Ichnos

An International Journal for Plant and Animal Traces

ISSN: 1042-0940 (Print) 1563-5236 (Online) Journal homepage: https://www.tandfonline.com/loi/gich20

Thunder lizard handstands: Manus-only sauropod

trackways from the Glen Rose Formation (Lower
Cretaceous, Kendall County, Texas)

James O. Farlow, Robert T. Bakker, Benjamin F. Dattilo, E. Everett Deschner,

Peter L. Falkingham, Crystal Harter, Richard Solis, David Temple & William
Ward

To cite this article: James O. Farlow, Robert T. Bakker, Benjamin F. Dattilo, E. Everett Deschner,
Peter L. Falkingham, Crystal Harter, Richard Solis, David Temple & William Ward (2019): Thunder
lizard handstands: Manus-only sauropod trackways from the Glen Rose Formation (Lower
Cretaceous, Kendall County, Texas), Ichnos, DOI: 10.1080/10420940.2019.1698424

To link to this article: https://doi.org/10.1080/10420940.2019.1698424

View supplementary material

Published online: 13 Dec 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=gich20
ICHNOS
https://doi.org/10.1080/10420940.2019.1698424

Thunder lizard handstands: Manus-only sauropod trackways from the Glen

Rose Formation (Lower Cretaceous, Kendall County, Texas)
James O. Farlowa, Robert T. Bakkerb, Benjamin F. Dattiloa, E. Everett Deschnerc, Peter L. Falkinghamd,
Crystal Hartere, Richard Solisc, David Templeb and William Wardc

a
Department of Biology, Purdue University Fort Wayne, Fort Wayne, IN; bHouston Museum of Natural Sciences, Houston, TX; cHeritage
Museum of the Texas Hill Country, Canyon Lake, TX; dSchool of Natural Sciences and Psychology, Liverpool John Moores University,
UK; ePurdue University Fort Wayne, Fort Wayne, IN

ABSTRACT KEYWORDS
Three parallel, manus-only sauropod trackways from the Coffee Hollow A-Male tracksite Sauropods; ichnology;
(Glen Rose Formation, Kendall County, Texas) were studied separately by researchers from manus-only trackways; Glen
the Heritage Museum of the Texas Hill Country and the Houston Museum of Natural Rose Formation
Sciences. Footprint and trackway measurements generally show good agreement between
the two groups’ data sets. Footprints appear to be shallowly impressed true tracks rather
than undertracks. One of the Coffee Hollow trackways shows marked asymmetry in the
lengths of paces that begin with the left as opposed to the right forefoot, and two of the
Coffee Hollow trackways are unusually broad. The Coffee Hollow trackways differ enough
from the manus portions of other Glen Rose Formation sauropod trackways to suggest that
they were made by a different kind of sauropod. Greater differential pressure exerted on
the substrate by the forefeet than the hindfeet probably explains the Coffee Hollow track-
ways, like other manus-only sauropod trackways, but the possibility that they indicate
unusual locomotion cannot at present be ruled out.

Introduction doubt made by a sauropod, but as I interpret them,

In early 1940, while searching for a suitable sauropod
trackway to display behind the American Museum of
Natural History’s Apatosaurus mount, fossil collector
Roland T. Bird briefly visited southern Texas. Late in
1938, Bird had found spectacular sauropod tracks in
the bed of the Paluxy River further north (Bird 1939,
1941, 1985), but he now hoped to avoid having to
deal with the Paluxy’s frequent floods (Bird 1985).
Although Bird did find some remarkable sauropod
ichnites in southern Texas (Bird 1944, 1954), he
ultimately had to deal with the temperamental Paluxy
to get the specimens he sought (Bird 1941, 1944,
1954). But while he was in southern Texas, Bird dis-
covered a kind of sauropod trackway whose interpret-
ation remains contentious.
Reported ‘elephant tracks’ on the Mayan Ranch in
Bandera County turned out to constitute an unusual
sauropod trail (Figure 1):
‘I saw them while we were still 100 feet away. A
double row of large, round circular prints. Without a
made by an individual while swimming. They were all
typical forefeet impressions as if the animal had just
been barely kicking bottom. (letter from Bird to
Barnum Brown, 8 February 1940)
Bird’s interpretation that Mayan Ranch sauropod
trackway had been made by a ‘swimming’ dinosaur
(Bird 1944, 1954) was uncontroversial at a time when
sauropods were thought to have been mainly aquatic
animals (Colbert 1951). A generation later, however,
the skeletal anatomy of sauropods was shown to be
more like that of fully terrestrial than amphibious
amniotes (Bakker 1971; Coombs 1975). Bird’s inter-
pretation of the Mayan Ranch sauropod trail was now
called into question, and an alternative hypothesis
offered: that manus-only or manus-dominated sauro-
pod trackways instead are due to differential impres-
sion of the forefeet and hindfeet during normal
quadrupedal locomotion, with manus and not pes
prints being registered in sediment layers beneath the
one on which the dinosaur actually walked (Lockley
and Rice 1990). Various authors have supported either

CONTACT James O. Farlow farlow@pfw.edu Department of Biology, Indiana-Purdue University, 2012 East Coliseum Boulevard, Fort Wayne,
IN 46805

Deceased.
Supplemental data for this article can be accessed at https://doi.org/10.1080/10420940.2019.1698424.
ß 2019 Informa UK Limited, trading as Taylor & Francis Group
2 J. O. FARLOW ET AL.

Figure 1. R. T. Bird’s Mayan Ranch ‘swimming’ sauropod trackway. (A–B) Photograph and map of the trackway as seen in 1940.
Most of the footprints were made by the left and right forefeet of the dinosaur; the single pes print noted by Bird is labelled. (C)
Cast (negative copy) of a right manus print from the trackway made by Farlow. Note the hoof-like shape of the print; anterior to
the top. Maximum print width ¼ 62 cm. (D) Bird’s cartoon of how he interpreted the progression of the trackmaker. Panels A, B,
and D from Bird (1944), and reproduced courtesy of the American Museum of Natural History.

the swimming or the under-tracking hypothesis, but
most authors have favoured the latter explanation (cf.
Ishigaki 1988, 1989; Lockley and dos Santos 1993;
Santos et al. 1994; Lee and Huh 2002; Henderson
2004; Vila, Oms, and Galobart 2005; Lee and Lee
2006; Hwang et al. 2008; Marty 2008; Ishigaki and
Matsumoto 2009; Milner and Lockley 2016; Xing
et al. 2016b). Computer simulation studies have also
demonstrated that due to differential underfoot pres-
sures, this effect may even occur on the tracking sur-
face, and not just in undertracks (Falkingham et al.,
2011a, Falkingham, Bates, and Mannion 2012).
Vila, Oms, and Galobart (2005) and Ishigaki and
Matsumoto (2009) presented a new approach to inter-
preting the origin of manus-only sauropod trackways:
If the trackway pattern of manus prints in manus-
only sauropod trackways does not differ from the
arrangement of manus prints in ‘standard’ (manus
plus pes print) trackways, the null hypothesis should
be that manus-only trackways were not made by dino-
saurs moving in any unusual manner—that is, that
the absence of pes prints is a formational or preserva-
tional artefact. In contrast, if the arrangement of
manus prints in manus-only trackways differs sub-
stantially from that in typical trackways, then perhaps
the trackmakers were indeed doing something out of
the ordinary when making the trackways.
In early 2007, the late Wann Langston of the
University of Texas received a telephone call from Ms.
Gail Flach, who owned property between Comfort
and Sisterdale in Kendall County, Texas. Ms. Flach
operated a limestone quarry (Coffee Hollow
Limestone) on her property, selling rock slabs to local
builders and contractors. She described the discovery

Figure 2. Coffee Hollow A-Male tracksite map. Inset shows location of the site in the state of Texas.
of a series of what she thought were dinosaur tracks,
and asked if Langston would be interested in coming
to look at them. Langston contacted the Heritage
Museum of the Texas Hill Country (HMTHC), whose
volunteers agreed to work on the site.
At the same time, Ms. Flach also contacted Bakker
at the Houston Museum of Natural Science (HMNS),
asking if his institution would be interested in her dis-
covery. As a result, two groups independently worked
at the Coffee Hollow A-Male tracksite (the second
part of the site name honouring an award-winning
purebred Mammoth Donkey at the ranch on which
the quarry is located), each unaware of the other’s
activities. Realisation that they were each working on
the same locality came when Bakker contacted Farlow,
ICHNOS 3
who had been advising the HMTHC, to see if Farlow
wanted to join the HMNS effort. With that the two
groups agreed to pool their efforts in a description of
the site.
In October 2007, under commercial pressure to fill
orders for her premium limestone, Ms. Flach had her
workers break up the track layer to sell slabs to local
builders. Fortunately, by this time the site had
been documented.
The Coffee Hollow A-Male dinosaur tracksite pre-
serves manus-only sauropod trackways like those dis-
covered by R.T. Bird in Bandera County, and by other
workers at other sauropod tracksites around the
world. The Coffee Hollow A-Male tracksite therefore
provides an opportunity to revisit the contentious
4 J. O. FARLOW ET AL.

Figure 3. Ground-level views of the three Coffee Hollow A-Male trackways. (A) The HMTHC field crew at work, cleaning and docu-
menting the site. Individual footprints are labelled following HMTHC naming conventions. (B) Detail of trackway A (Right trackway
of HMNS). (C) Detail of trackway C (Left trackway of HMNS).

issue of how manus-dominated sauropod trackways
were made. Although we will not definitively answer
the question of how the Coffee Hollow trackways
were made, we will discuss the kinds of trackway evi-
dence that would allow such a determination.
Geologic setting
The Coffee Hollow A-Male dinosaur tracksite con-
sisted of three trackways in a level limestone layer
exposed by quarrying operations (Figures 2–5).
Limestone from this layer is much in demand as a
building stone, and decorative features such as fossils,
ripple marks and animal burrows are also sought
after. The track surface had been exposed by remov-
ing overburden of up to twenty feet of fossiliferous,
marly limestone. It was bounded on the west by a nat-
ural slope, on the northeast by quarried area and fill,
and on the southeast and south by unquarried over-
burden. The extent of the trackways to the southeast
was not determined, and probably had not
been reached.
The track layer (commercially known as the
Sisterdale Cream Stone) lies about a metre below the
 ICHNOS 5

Figure 4. Overall tracksite images. (A) Digital model of the tracksite created from digital photographs taken by the HMTHC field
crew. Colour and contrast are artificially altered to enhance the distinctiveness of the footprints. Some footprints have a ghostly
‘double-strike’ configuration due to imperfections in the model. (B) Image of the tracksite made from photographs taken by the
HMNS field crew using a drone; scale bar in figure applies specifically to this panel. In both images footprints are labelled using
HMTHC naming conventions.

base of the Salenia texana Zone (Whitney 1952; Scott
et al. 2007; Ward and Ward 2007), and about 4 m
below the ‘Corbula’ Marker Bed at the top of the
Lower Glen Rose Member of the Lower Cretaceous
Glen Rose Formation. This is about the same strati-
graphic position as very large sauropod tracks exposed
in the bed of the Blanco River about three miles west
of Blanco, Texas (Pittman 1989; Scott et al. 2007;
Ward and Ward 2007). The Sisterdale Cream Stone
is a wackestone-packstone containing burrows and
ostracodes (Figure 5). In addition, some distinctive
vertebrate skeletal fossils were found during quarry-
ing operations, including fishes and a small crocody-
liform (now in the collection of the Heritage
Museum of the Texas Hill Country, Canyon Lake,
Texas). The exact stratigraphic position at which
these vertebrate fossils were found is uncertain, but
may have been just above the tracklayer, close to the
Salenia texana zone. A distinctive joint pattern was
oriented N20 W; one major joint extended through
the central portion of the site (cutting across print
A83 of the Right trackway), and minor joints
spread throughout.
Methods
Because both HMTHC and HMNS field crews inde-
pendently worked the Coffee Hollow A-Male tracksite,
a description of the procedures followed by both
groups is necessary.
HMTHC protocols
The three sauropod trackways were labelled Trails A,
B, and C, going from north to south across the site
(Figures 2–4); the trackways are essentially parallel,
with all three dinosaurs moving in a northwesterly
direction, animal A being the rightmost of the three,
6 J. O. FARLOW ET AL.
Figure 5. Stratigraphic section of the Glen Rose Formation as exposed at the Coffee Hollow A-Male tracksite.
and animal C the leftmost. Because the direction of
travel of the trackmakers was towards a naturally
occurring slope, and additional tracks were still being
uncovered by quarrying operations further ‘up-trail’
(in the direction from which the dinosaurs had come)
in the three trackways, footprints were numbered
arbitrarily, starting with footprint 100 at the outcrop
and numbered sequentially backward (up-trail). For
each of the trackways, footprints of the right side
were assigned odd numbers, and prints of the left side
even numbers.
Trimble Business Center software was used to link
GPS equipment (VRS) to a Trimble 5600 robotic total
station enabling rapid acquisition of targets with near-
centimeter accuracy. These points are referenced to
the North American Datum of 1983 (NAD 83) as the
horizontal control datum for the United States, and
the North American Vertical Datum of 1988 (NAVD
88) as the vertical control datum of optometric height
established for vertical surveying in the U.S. With this
system, points were acquired around the perimeter of
each mapped footprint with a maximum of a few cen-
timetres between each point. At least 10 such points
were taken for each print; for many prints, there were
30 or more. In addition to the footprint perimeters,
the boundary of the track surface (bluff edge, quarried
edge and unexcavated boundary) was mapped, as well
as several of the major joints, giving the orientation of
these features. Back in the office, digital photographs
were oriented, scaled and superimposed over the
mapped track perimeters, confirming the accuracy of
the mapping.
Prints 72 through 99 of the A trackway were
mapped, as were prints 87–97 of the B trackway, and

prints 74–99 of the C trackways. Other footprints
were seen, but were broken up before they could be
mapped: prints 58–71 of the A trackway, prints 78–86
of the B trackway, and prints 68–73 of the C track-
way. In addition, tracks A73, A98, B85, B86, C75 and
C77, though within the mapped portions of the trails,
were either too poorly preserved or poorly exposed to
map. Other footprints were reported to have been
uncovered after the HMTHC crew had finished its
work, such that there may have been more than 100
individual footprints at the site.
Measurements of individual footprints and of
trackways were directly made with a tape measure in
inches and feet (later converted to metric units).
Footprint lengths and widths were measured (to the
nearest half-inch). Oblique paces and strides were
measured (to the nearest inch) from the most anterior
location along the front margin of each print, as was
done by Farlow, Pittman, and Hawthorne (1989) for
other Glen Rose Formation sauropod trackways. For
comparison, oblique pace lengths were later made
from footprint centres using the site map.
Measurements of inner and outer trackway width
were made at the positions of each print along the
length of the trackway, as defined by the innermost
and outermost edge, respectively, of the print. Pace
angulations were not measured directly, but calculated
from pace and stride measurements using the law of
cosines; the width of the angulation pattern (Marty
2008) was calculated from paces, strides, and the pace
angulation using the law of sines (Farlow, Pittman,
and Hawthorne 1989).
Footprint rotations (positive values indicate out-
ward rotation with respect to the trackmaker’s direc-
tion of travel, and negative values inward rotation)
were not measured in the field, but a few estimates
were later made from the trackway map (Figure 2). A
line was run through the middle of each of trackways
A and C, and the angle formed between the antero-
posterior axes of some of the better-registered foot-
prints in each trackway and the line through the
middle of the trackway measured.
Numerous digital photographs of trackways and
individual footprints were taken, from a variety of
angles, during all stages of field work. Photographs of
individual footprints were taken from a stepladder at
a height of about 71=2 feet (2.3 metres) above the
ground, with an American yardstick marked off in
inches to indicate the scale. These photographs were
taken from as nearly directly overhead as possible, but
most of them were a bit oblique. The brightness and
contrast of photographs were altered to make the
ICHNOS 7
prints stand out from the surrounding rock. Many of
the trackway and tracksite photographs were later
processed using COLMAP (Sch€ onberger and Frahm
2016; Sch€ onberger et al. 2016) to build a post-hoc
photogrammetric digital model (Falkingham 2012;
Falkingham, Bates, and Farlow 2014, Falkingham
et al. 2018) of the tracksite (Figure 3(A)). Because the
photographs were not originally taken with photo-
grammetry in mind, overlap was inconsistent, and
objects such as scale bars and equipment moved
around between images. This meant that 3 D detail
and resolution of the tracksite model was relatively
low, and useful primarily as a 2 D map (Figure 4(A)).
HMNS protocols
The three trackways were identified in terms of their
position, compared with the direction of the dino-
saurs’ trackway. The HMNS Left Trackway is
HMTHC trackway C, the HMNS middle trackway is
HMTHC trackway B, and the HMNS Right Trackway
is HMTHC trackway A.
Footprint width was measured directly on each
print to the nearest millimetre. A reference line was
run through the trackway along its overall direction of
travel, and marked at four-foot intervals. Each of these
four-foot increments was marked on graph paper
(Figure 6), and the position of each footprint drawn
on the paper. The perpendicular distance from the
innermost rear edge of the footprint to the reference
line was measured, as was the angle formed by a line
segment running across the back margin of the foot-
print with the reference line.
Oblique paces and strides were measured indirectly,
from the trackway diagrams. To make these data as
comparable to the HMTHC data as possible, paces
and strides were measured from the anterior edges of
the footprints as drawn on the trackway diagrams.
Measurements made on the trackway diagram were
converted to real world values by comparison with the
distance on the diagram corresponding to the four-
foot reference line increment marks. Pace angulations
were measured on the trackway diagrams.
Two measurements of inner trackway width were
made, the first of these a direct measurement. The
distance from the inner rear edge of one footprint to
the reference line was added to the distance from the
inner rear edge of the immediately preceding contra-
lateral footprint to the reference line, and this sum
was taken to indicate the inner trackway width at the
position of the more down-trail (in the direction
towards which the dinosaur was moving) footprint.
8 J. O. FARLOW ET AL.
Figure 6. Diagram (redrawn from field notes) illustrating how
HMNS footprint and trackway measurements were made. To
facilitate comparisons, the HMTHC labels for footprints are
added to the diagram. Numbers along the left margin (32–56)
indicate some of the four-foot increments marked off on the
field diagram. Individual footprint widths are indicated (in mm)
above each print; thus the width of print C78 is 630 mm. A
line is run through the trackway along its direction of travel,
and its orientation relative to magnetic north indicated (bot-
tom of figure; in this case the direction of travel is 287 . The
measured distance from the inner rear edge of the footprint
to the reference line is indicated above a short line segment
running from the edge of the print to the reference line; for
footprint C81, this distance is 615 mm. A line segment running
across the rear margin of the print forms an angle with the
reference line, and is drawn on the diagram next to the foot-
print; for print C78, the angle is 73 . The bearing of the foot-
print relative to the dinosaur’s direction of travel will be
perpendicular to the line running across the print’s
rear margin.
Inner trackway widths were also measured from
the trackway diagrams by drawing line segments con-
necting the innermost edges of two successive ipsilat-
eral footprints, and then measuring the perpendicular
distance from that line segment to the innermost edge
of the intervening contralateral footprint, and finally
converting this measurement to its real-world value.
Thus there could be two fewer measurements of inner
trackway width than the number of footprints in alter-
nating left-right sequence along the length of the
trackway. Outer trackway width was measured in an
analogous manner, except that, of course, it was meas-
ured using the outer rather than the inner edges
of footprints.
Footprint rotation with respect to the dinosaur’s
direction of travel was calculated in a two-step oper-
ation. The compass bearing of a footprint is perpen-
dicular to the angle formed by the line segment
running across the back margin of the footprint and
the reference line indicating the dinosaur’s travel dir-
ection (Figure 6). For left footprints, this bearing is
calculated as: direction of travel (in degrees) þ angle
formed by the line segment running across the back
margin of the footprint, minus 90 . For left footprint
C78 (Figure 6), this angle will be 287 þ 73 minus
90 , or 270 . This bearing is 17 less than the dino-
saur’s overall direction of travel, and so this footprint
is rotated outward (away from the overall direction of
travel) by 17 . For right footprints, the individual
print bearing is calculated as direction of travel minus
angle formed by the line segment running across the
back margin of the footprint þ 90 .
Overlapping photographs of the tracksite taken
from overhead, using a drone, were combined to cre-
ate a composite tracksite photograph that can be com-
pared with the HMTHC tracksite digital model
(Figure 4(B)).
Negative latex copies (casts) of some of the prints
were made. These were made after cleaning out loose
sediment covering the prints, but without any attempt
to remove any solidly adhering possible footprint fill
material. These latex peels were later photographed by
Farlow with photogrammetry in mind, and because of
this the digital models created from the photographs
by Falkingham have much greater resolution than the
models for the full site, and contain reliable informa-
tion in all three dimensions.
Track layer and track fill sampling
The HMTHC team collected rock samples associated
with three of the footprints. Footprint A99 from the
 ICHNOS 9

Figure 7. Cross-sections and fill associated with footprints. (A–C) Manus print A99 (Right trackway). A. The broken footprint in situ.
(B) Oblique view of the footprint, seen from the top left of the print as displayed in panel A, showing part of a section across the
footprint. Note slight displacement rim that slopes down into the bowl-shaped depression of the footprint proper. (C) Pieces of
the cross-section across the footprint. (D) Two facing sides of a polished section through a rock piece interpreted as overtrack fill
of A97 or a nearby print. The lower view is printed with left and right reversed, so that its face is directly comparable with that of
its counterpart in the upper view. This piece sat atop the material interpreted as the track layer; if this interpretation of the track
layer is correct, the base of the fill would be at the bottom of the two sections illustrated here. (E, F). Fill of manus print (either
C83 or a nearby print in the Left trackway). (E) Cut piece through the fill (the label on the slab identifying this piece as coming
from print C70 is incorrect). The bottom of the piece as illustrated here sat atop the footprint as exposed in the quarry. (F) View
of the underside of the fill piece, showing the location of the cut piece from panel E. (G) Broken fill piece partially covering print
A86, with ripples atop the fill.
10 J. O. FARLOW ET AL.
Right trackway (Figure 2) was at the broken edge of
the track exposure, making it possible to collect a sec-
tion across the print (Figure 7(A)–(C)). Rock material
filling footprints was collected from print A97 or a
nearby print (Figures 7(D)) and from a print in the C
(Left) trackway, print C83 or nearby (Figures 7(E, F)).
We have little direct information about the material
comprising footprints in the B (Middle) Trackway.
Cut surfaces across prints and/or fill were polished to
facilitate interpretation of gross features, and thin sec-
tions cut for petrological analysis of microscopic fea-
tures (Figures 8 and 9).
Comparative trackway data (supplementary online
material Table 1)
We scoured the dinosaur trackway literature for meas-
urements of the manus portion of sauropod track-
ways, including ‘standard’ trackways composed of
both manus and pes prints, trackways that were expli-
citly characterized by the authors as ‘manus-domin-
ant’ (manus and pes prints both present, but manus
prints more deeply or clearly impressed than pes
prints), and manus print-only trackways. We admit
that the distinction between ‘standard’ and ‘manus
dominant’ trackways is somewhat arbitrary. Some of
what we characterize as standard trackways may well
have been manus-dominant, but unless the authors of
the studies so designated them, they are not labelled
as manus-dominant by us. We will distinguish
between standard and explicitly identified manus-
dominant trackways in many of our graphs, but for
many data analyses we will group the two trackway
categories together.
Data are variably reported in the literature.
Measurements of manus width are commonly
reported (and will be used as the chief proxy for
trackmaker size in this study), as are pace lengths and
stride lengths, and pace angulations, but the details of
how measurements were made are seldom reported,
and may vary among studies. For example, Farlow,
Pittman, and Hawthorne (1989) measured manus and
pace lengths of the classic Glen Rose Formation
Brontopodus trackways using as reference points the
front margins of the manus prints, rather than the
print centres, because in some Paluxy River trackways
manus prints are squashed from the rear by sediment
squeezed forward during impression of the pes. In
most published measurements of sauropod manus
step lengths, however, print centres were probably
used as the reference points. Trackway widths may be
expressed as inner or outer trackway widths, and/or
the width of the angulation pattern (cf. Marty 2008).
For comparisons among trackways, we used the
means of published values of parameters for each
trackway, if more than one measurement of the par-
ameter was reported by the authors; if only a single
measurement was available, this was the value
employed. In many cases values of one or more
parameters were not reported, but could be estimated
from the author’s trackway diagrams. We will make
quantitative comparisons of sauropod trackways using
all of these kinds of data extracted from the literature,
but our results should be taken with caution due to
the limitations of those data. Some trackway parame-
ters are geometrically related (cf. Farlow, Robinson,
Turner, et al. 2018b for crocodile trackways), such
that values that authors did publish can serve as prox-
ies for parameters that were not reported.
Forelimb length vs. Manus size in sauropods
To estimate the water depth necessary to float a
sauropod while its forefeet maintained contact with
the bottom, and assuming that water would have
reached the animal’s shoulder, we compared the
manus width and forelimb length of mounted sauro-
pod skeletons. Manus width and forelimb length were
measured from digital models collected during a pre-
vious study (Bates et al. 2016) via laser scanning or
photogrammetry. Measurements were made using
Autodesk Maya 2017. Forelimb length was measured
as the combined segment lengths of humerus, radius,
and manus; this undoubtedly underestimates the
actual length. Manus width was measured as the max-
imum distance across the manus in a medial-lateral
direction. Measuring forelimb length as summed seg-
ment lengths makes the result resistant to pose.
However, manus width is more dependent on the way
in which the skeleton (manus) is reconstructed.
Compounding this issue, distal portions of the auto-
podia were generally captured least well by both laser
scanning and photogrammetry, and often contained
extensive reconstructions of missing autopodial ele-
ments. These data therefore represent rough
approximations.
Results
Track layer and overtrack fill lithology
In cross section (Figure 7(A)–(C)), the manus prints
are a few centimetres deep, with at least some prints
showing an elevated displacement rim (Marty,
 ICHNOS 11

Figure 8. Thin sections through pieces of the track layer and overtrack fill. Scale bars ¼ 1 cm. (A, B). Several sections (labelled
with red letters) through the left portion of the slab cut across manus print A99 shown in Figure 7C; the sections were cut
through the underside of the footprint as illustrated in Figure 7C. (A) The base of the footprint occurs along the marginal edge of
section A3, and the top edges of sections B2 and B4. Details of portions of the columns labelled 1 - 3 appear in Figure 9. (B)
Interpretive drawing showing the four microstratigraphic layer units observed vertically across the sections. Note especially the
compression of unit 4, relative to underlying layers, beneath the footprint floor as opposed to outside the print (column 1 vs. col-
umns 2 and 3 in panel A); also see Figure 9A–C. (C, D). Thin sections (C) and interpretive drawing (D) of the seven microstrati-
graphic units observed vertically across the sections through the overtrack fill of manus print A97 (or a nearby print in the
trackway); compare with Figure 7D, which shows the two opposing faces of the same surface. An enlarged image of a portion of
the column labelled ‘Detail’ is shown in Figure 9.

Falkingham, and Richter 2016) around the print mar-
gin. Because the upper surface of the layer in which
the tracks are observed shows greater vertical deform-
ation associated with the prints (Figures 7(B), (C),
and 8(A)) than the material overlying it (Figures 7(D)
and 8(C)), we interpret the manus prints as true
tracks rather than undertracks (Marty, Falkingham,
and Richter 2016; cf. Sanz et al. 2016). Furthermore,
the fill layer overlying some prints shows ripple marks
(Figure 7G), which would have been squashed had the
fill layer itself been the one on which the dino-
saur trod.
What we therefore interpret as the track layer
(Figures 7(C), 8(A), and 9(A)–(C)) consists of four
fairly similar, repetitive sedimentary units, each of
which fines upward from fine-grained packstone to
mudstone. Most of the grains are ostracodes, dasy-
clads, and foraminiferan tests. Scattered through the
12 J. O. FARLOW ET AL.

Figure 9. Details of thin sections through cross sections of a footprint and fill. Scale bars ¼ 2 cm. (A–C). Portions of columns 1
(beneath the footprint), 2 (through and beneath the displacement rim), and 3 (outside the footprint), respectively, from Figure 8A.
Numbers along the margins of the slices label four sequential layer units (Figure 8B), with tic marks indicating boundaries between
the units. (D). Portion of the section through footprint fill labelled ‘Detail’ in Figure 8C. Numbers along the margin of the slice label
the seven sequential layer units through the fill (Figure 8D), with tic marks indicating boundaries between the units.

track layer are sparse, subangular to subrounded
quartz sand grains whose diameters range 0.3–0.8 mm,
with a concentration of diameters at 0.4–0.5 mm.
Diagenetic mud clots of 0.2–0.4 mm diameter are dis-
tributed through the layer, along with birdseye struc-
tures of comparable size. In addition to the fossils
already listed, there are small (0.5–3 mm across) gas-
tropods and also small (up to 7 mm across), thin
(0.05–0.1 mm thick) fragments of possible crustacean
or limulid carapace; the abundance and diversity of
body fossils decreases vertically across the layer. The
layer is marked by thin (< 1 cm) burrows and occa-
sional (reworked?) worm (?) tubes. The uppermost
portion of the track layer (unit 4) is markedly
deformed by the dinosaur footprint, being squashed
beneath the track and pushed up away from it.
The overall environmental interpretation of the
track layer is of a subaqueous (indicated by the graded
beds), shallow-water situation, very close to shore
(source of the sand grains), with a low-diversity,
sparse (stressed?) benthic fauna. The clotted structures
and birdseyes suggest algal or microbial growth; the
presence of dasyclads suggests a closed lagoon (C.
Meyer, personal communication).
The material filling footprints has a more com-
plicated microstructure than the material of the
track layer (Figures 7(D), 8(C, D), and 9(D)). The
basal sublayer (unit 1) may have proceeded higher
sublayers by a prolonged interval of time, because
those upper layers show onlap over it. The very
base of unit 1 is a foraminiferan/ostracode
packstone, the forams of which are similar to those
of the track layer. The remainder of unit 1 is a
mudstone or wackestone with vertical birdseye
structures separating possible microstromatolitic
columns 2 and 3 mm across.
Fossils are rare or absent in the track fill laminae
above unit 1. Unit 2 shows faint microlaminations
(0.1 and 0.2 mm thick) composed of angular, discrete
mudstone fragments or lithoclasts. There are scattered,
equant fenestrae (0.2–0.5 mm) at a uniform depth
along the upper surface of the unit. Unit 3 consists of
thin lamellae like those of unit 2, with fenestrae of
0.1–0.5 mm diameter lined up along the base of the
unit. The remaining units (4–7) appear to be traction
deposits of micro-lithoclasts. These are monomictic,
well-rounded, nearly spherical (length/width ratio up
to 1.5) bits of lime mud of 0.1–0.3 mm diameter in a
packstone or grainstone matrix. Unit 6 shows moldic
porosity, with nearly spherical, well-rounded voids
that are 0.1–0.5 mm across. There is no bioturbation
of the overtrack fill.
 ICHNOS 13

Table 1. Measurements of Coffee Hollow A-Male sauropod trackways. Linear dimensions in cm, angles in degrees.
HMTHC ¼ measurements made by Heritage Museum of the Texas Hill Country team; HMNS ¼ measurements made by Houston
Museum of Natural Science team. Direct ¼ measurements made in the field, directly on the trackway, or (for angular measure-
ments) calculated from such direct measurements. Diagram ¼ measurements made from scaled trackway diagram.
Map ¼ measurements made from tracksite map.
Trackway Parameter Mean (Median) Range N
Coffee Hollow A-Male A (Right Trackway) Manus length (HMTHC direct) 70.4 (71.1) 64–76 24
Manus width (HMTHC direct) 70.0 (69.9) 61–76 24
Manus width (HMNS direct) 71.4 (70.0) 66–79 25
Manus rotation (HMTHC map) 33.9 (30.0) 22–47 7
Manus rotation (HMNS direct) 29.4 (33.0) 13–47 21
Manus pace (Print front edge; HMTHC direct) 246.9 (236.2) 198–200 21
Manus pace (Print centre; HMTHC map) 236.9 (253.8) 202–278 23
Manus pace (Print front edge; HMNS diagram) 234.1 (227.3) 176–298 17
Manus stride (Print front edge; HMTHC direct) 254.0 (266.7) 170–292 19
Manus stride (Print front edge; HMNS diagram) 257.4 (254.0) 181–401 15
Manus trackway inner Width (HMTHC direct) 126.9 (127.0) 114–145 22
Manus trackway inner Width (HMNS direct) 129.1 (132.0) 56–157 18
Manus trackway inner Width (HMNS diagram) 96.6 (95.0) 78–116 15
Manus trackway outer width (HMTHC direct) 275.8 (274.3) 246–295 22
Manus trackway outer width (HMNS diagram) 209.6 (211.0) 196–122 15
Manus width of angulation pattern (HMTHC direct) 206.2 (208.3) 178–234 17
Manus pace angulation (HMTHC direct) 61.4 (64.0) 39–76 17
Manus pace angulation (HMNS diagram) 64.3 (63.0) 47–94 15
Coffee Hollow A-Male B (Middle Trackway) Manus length (HMTHC direct) 52.9 (53.3) 44–58 11
Manus width (HMTHC direct) 51.6 (53.3) 46–56 11
Manus width (HMNS direct) 53.5 (52.0) 41–64 8
Manus pace (print front edge; HMTHC direct) 288.0 (287.0) 279–302 10
Manus pace (Print centre; HMTHC map) 278.0 (278.0) 256–309 10
Manus pace (Print Front edge; HMNS diagram) 290.6 (273.9) 261–335 7
Manus stride (HMTHC direct) 375.7 (369.6) 356–419 10
Manus stride (HMNS diagram) 343.9 (352.2) 278–419 6
Manus trackway Inner Width (HMTHC direct) 170.9 (170.2) 157–188 11
Manus trackway Inner Width (HMNS direct) 216.1 (212.0) 192–247 7
Manus trackway Inner Width (HMNS diagram) 179.1 (178.1) 164–194 6
Manus trackway outer width (HMTHC direct) 260.0 (256.5) 254–274 11
Manus trackway outer width (HMNS diagram) 260.9 (262.4) 247–278 6
Manus width of angulation pattern (HMTHC direct) 217.3 (221.0) 196–236 9
Manus pace angulation (HMTHC direct) 81.9 (80.0) 74–94 9
Manus pace angulation (HMNS diagram) 71.5 (71.5) 63–80 6
Coffee Hollow A-Male C (Left Trackway) Manus length (HMTHC direct) 65.0 (64.8) 58–70 16
Manus width (HMTHC direct) 67.0 (66.0) 61–86 16
Manus width (HMNS direct) 64.9 (64.0) 56–73 13
Manus rotation (HMTHC map) 21.5 (21.5) 7–33 6
Manus rotation (HMNS direct) 23.3 (23.0) 11–43 12
Manus pace (Print front edge; HMTHC direct) 194.8 (189.2) 152–226 16
Manus pace (Print centre; HMTHC map) 181.6 (182.7) 155–211 21
Manus pace (Print front edge; HMNS diagram) 187.7 (191.7) 157–202 11
Manus stride (HMTHC direct) 262.8 (259.1) 251–282 15
Manus stride (HMNS diagram) 263.5 (264.9) 230–290 10
Manus trackway inner width (HMTHC direct) 74.9 (74.9) 64–89 10
Manus trackway inner width (HMNS direct) 67.5 (66.5) 52–85 6
Manus trackway inner width (HMNS diagram) 53.9 (50.7) 45–68 10
Manus trackway outer width (HMTHC direct) 195.6 (191.8) 185–208 10
Manus trackway outer width (HMNS diagram) 162.2 (160.7) 147–182 10
Manus width of angulation pattern (HMTHC direct) 138.5 (137.2) 127–152 15
Manus pace angulation (HMTHC direct) 86.1 (84.0) 80–95 15
Manus pace angulation (HMNS diagram) 88.8 (90.5) 72–95 10

The surface of the fill material of some prints
shows a wrinkled texture suggestive of a microbial
mat (Figures 10(A) and 11(A)), or ripple marks
(Figure 7(G)). Overall the track fill material suggests
deposition on a tidal flat, or even a supratidal situ-
ation that was only occasionally flooded, in either case
probably with an algal mat cover (cf. Kvale et al.
2001; Paik et al. 2017; Marty, Strasser, and Meyer
2009; Diedrich 2012; Carvalho, Borghi, and Leonardi
2013; Alcala et al. 2014; Cariou et al. 2014; Dai et al.
2015; Campos-Soto et al. 2017; C onsole-Gonella et al.
2017; Paik et al. 2017). Similar overtrack fills are
known from tridactyl footprints from the Glen Rose
Formation of southern Texas (Farlow et al. 2006).
Because we have no information about the material
underlying trackway B, our inferences about whether
the Coffee Hollow trackways are true tracks as
opposed to undertracks apply mainly to trackways A
and C. Indeed, the footprints in trackway B are par-
ticularly faint, so of the three trackways this one is the
14 J. O. FARLOW ET AL.
Figure 10. HMTHC B (HMNS Middle) trackway footprints. The brightness and contrast of photographs have been manipulated to
increase the clarity of print outlines. Prints with an odd number are rights, and prints with an even number are lefts. See Figure 2
for locations of prints in the trackway. (A) B88, (B) B89, (C) B90, (D) B93, (E) B94, and (F) B96 overlapping print C97.
best candidate for being composed of underprints, but
we have no other reason for thinking this to be the
case. The wrinkled texture of the surface of some track-
way B prints (Figure 10(A)), as already noted, suggests
the presence of a fill layer covering true tracks.
Footprint morphology and preservation
As with many other sauropod trackways (Castanera
et al. 2016b), manus prints of the three Coffee Hollow
A-Male trackways present as sub-circular, elliptical,
crescent-shaped, or hoof-shaped depressions (Figures
7(A), (F), (G), 10–13). Like manus prints of the Mayan
Ranch trackway (Figure 1), they are shallowly
impressed (only a few to several cm maximum depths),
but in some cases the shallowness is probably due in
part to fill covering the footprint. Print outlines are par-
ticularly faint in the Middle (B) trackway (Figure 10).
If the size of the prints as preserved is a true indi-
cation of the size of the manus, these three dinosaurs
were among the bigger sauropod trackmakers known,
both from the Glen Rose Formation and from dino-
saur tracksites more generally (Figures 14–16). The
Left (C) and Right (A) trackways are similar in manus
print width, and about 25–35% larger than manus
prints of the Middle (B) trackway. If the prints as pre-
served were undertracks, it is possible that their
dimensions could exaggerate the sizes of the autopo-
dia that made them (cf. Farlow et al. 2006; Milan and
Bromley 2006; Marty 2008; Jackson, Whyte, and
Romano 2009, Castanera et al. 2012a; Thulborn 2012;
Xing et al. 2015a; Brusatte et al. 2016). As already dis-
cussed, however, the preservation of the prints,
particularly of trackways A and C, suggests that they
are true tracks rather than undertracks.
In well-registered sauropod manus or pes prints
the sole shows impressions or wrinkles indicative of
sole pads (cf. Farlow, Pittman, and Hawthorne 1989;
Dalla Vecchia and Tarlao 2000; Milan, Christiansen,
and Mateus 2005; Platt and Hasiotis 2006; Mateus and
Milan 2010; Huh et al. 2003; Romano and Whyte
2012; Castanera et al. 2016b; Hall, Fragomeni, and
Fowler 2016). Occasionally the skin texture of the sole
is recorded as an abutting polygonal pattern (Lockley
et al. 1998, 2008; Currie, Badamgarav, and Koppelhus
2003; Platt and Hasiotis 2006; Kim et al. 2010; Mateus
and Milan 2010; Navarette et al. 2014; Pi~ nuela Suarez
2015; Castanera et al. 2016a; Fondevilla et al. 2017;
Paik et al. 2017).
There is no indication of such sole scalation in the
Coffee Hollow manus prints as preserved (Figs. 10-
13). At a larger scale, however, the surface of the latex
peels of some of the better prints (Figure 13) shows a
rough texture reminiscent of that which occurs on the
sole of the hands and feet of elephants (cf. Hall,
Fragomeni, and Fowler 2016), and/or as ‘pull-up’ fea-
tures created by suction as the foot is withdrawn from
the sediment (Kvale et al. 2001). However, to the
extent that the moulded footprints had firmly adher-
ing overtrack fill, the texture of the peels reflects the
surface of the fill material rather than the soles of the
trackmakers’ forefeet.
If the Coffee Hollow footprints are true tracks,
then the absence of pes prints cannot be due to differ-
ential preservation of manus and pes prints as under-
tracks. However, differential pressure might result in
 ICHNOS 15

Figure 11. HMTHC C (HMNS Left) trackway footprints. The brightness and contrast of photographs have been manipulated to
increase the clarity of print outlines. Prints with an odd number are rights, and prints with an even number are lefts. See Figure 2
for locations of prints in the trackway. (A) C84, (B) C85, (C) C86, (D) C87, (E) C88, (F). 89, (G) C90, (H) C91, (I) C93, (J) C94, (K). C96,
(L) C98 and (M) C99.

manus-only prints even if the tracks were true tracks,
if there was more pressure loading on the forefoot
than on the hindfoot (Falkingham et al., 2011a). A
potential problem with the hypothesis of differential
loading as a factor in creating Glen Rose Formation
sauropod manus-only or manus-dominant trackways,
however, is that to date wherever else sauropod track-
ways showing both forefoot and hindfoot impressions
occur in this unit, the pes prints are always as deeply
impressed as, or more deeply impressed than, manus
prints (Figure 17; also see Farlow, Pittman, and
Hawthorne 1989, Farlow et al. 2015; Farlow 1992: figs.
10, 11).
On the other hand, support for the differential
pressure hypothesis comes from the CertainTeed
Gypsum mine (previously known as the Briar Site,
Nashville, Arkansas), in the De Queen Formation, a
sedimentary unit correlative to the Glen Rose
Formation. Long sauropod trackways were first
reported from this locality by Pittman and Gillette
(1989). More recently, Platt et al. (2018) described
theropod trackways and sauropod footprints from a
different part of the same quarry. Dinosaur footprints
at this site were interpreted as true tracks. In most
sauropod trackways from the Briar site, manus and
pes are about equally deeply impressed, but at one
16 J. O. FARLOW ET AL.

Figure 12. HMTHC A (HMNS Right) trackway footprints. The brightness and contrast of photographs have been manipulated to
increase the clarity of print outlines. Prints with an odd number are rights, and prints with an even number are lefts. See Figure 2
for locations of prints in the trackway. (A) A75, (B) A76, (C) A77, (D) A78, (E) A79, (F) A80, (G) A81, (H) A82, (I–J). A83, (I) Footprint
in situ. (J) Latex negative copy (cast) of the print; note large joint and solution feature cutting across the print, (K) A84, (L). A85,
(M) A87, (N) A88, (O) A89 and (P) A97.

portion of the site, where the substrate at the time of
track formation is thought to have been firm (Platt
et al. 2018), manus prints are slightly deeper than
associated pes prints (B.F. Platt, personal communica-
tion). Furthermore, there are manus prints in the
same area of the quarry that are unaccompanied by
pes prints (Platt et al. 2018). The sauropod prints
from the Briar Site are morphologically similar to
those from the Glen Rose Formation (Farlow,
Pittman, and Hawthorne 1989; Pittman and Gillette
1989). Consequently differential pressure cannot be
excluded as a possible explanation for manus-only
sauropod trackways in the Glen Rose Formation
(Falkingham et al. 2011a).
Consistency of measurements
Summary measurements for the three sauropod track-
ways are presented in Table 1, and internal correlations
between different measurements of the same parameter
within the HMTHC or HMNS data sets, and between
the HMTHC and HMNS data sets, are presented in
Table 2 and Figures 18 and 19. Some individual meas-
urements differ markedly between the two data sets,
due to differences in measurement protocols between
the two teams, difficulties in measuring these some-
times faint prints, or both. Most mean and median val-
ues of footprint and trackway parameters, however, are
gratifyingly very close (within a few centimetres of lin-
ear measurements, and a few degrees of angular meas-
urements) between HMNS and HMTHC values of the
same measurement; exceptions include some measures
of trackway inner and outer width, paces, strides, and
pace angulations. It should be noted, however, that the
calculated means and medians are generally not based
on the same number of measurements for the HMNS
and HMTHC data sets, which may account for some
 ICHNOS 17

Figure 13. Negative copies (casts) of Coffee Hollow A-Male sauropod footprints. Note the irregular texture of what is either the
print sole or the surface of footprint fill. (A, B) Prints from one of the large sauropods (probably [Right] trackway A). (A) Maximum
diameter across the print ¼ 67 cm; maximum relief ¼ 7 cm. (B) Maximum diameter ¼ 61 cm; maximum relief ¼ 3-4 cm. (C, D).
Footprint from the middle (B) trackway. (C) Photograph of the print; maximum diameter ¼ 48 cm. (D). Digital model of track relief;
the images are rotated clockwise to the view in panel C. Scale bar ¼ 10 cm. Top row ¼ plan view, bottom row ¼ oblique view.
Warm colours in the heat maps indicate deeper parts of the footprint.

of the differences in average values between the two
data sets. Correlations between HMTHC and HMNS
measurements of the same parameter (Table 2; Figure
18) are generally good (Pearson’s r at least 0.65, with
p < 0.001); the lowest correlations are between meas-
urements of trackway outer width (r ¼ 0.559,
p ¼ 0.003), and between measurements of footprint
rotation (which correlation is not statistically signifi-
cant; r ¼ 0.225; p ¼ 0.532, but is based on a small sam-
ple size [N ¼ only 10]).
HMTHC measurements of paces and strides using
different reference points (print front margin vs. print
centre) have mean values (Table 1) that differ by less
than 10%, with step lengths measured from the front
edges of the prints being longer than those measured
from the print centres. The two different ways of
measuring trackway inner width using the HMNS
data set yield mean values that differ by 20–30%, with
the direct method yielding larger mean values than
the estimate based on trackway diagrams. Not too
much should be made of this discrepancy, however,
because the two sets of measurements were made very
differently. This highlights a warning for comparisons
of trackway measurements among studies: before con-
cluding that values differ between trackways described
by different authors, one should be sure that the
measurements were made the same way (Falkingham
2016). In any case, the different ways of measuring
paces in the HMTHC data set, and trackway inner
width in the HMNS data set, are both highly corre-
lated (Table 2).
Inter-relationships among trackway parameters
Some of the trackway parameters are geometrically
related (Figure 19). After removing the effects of
trackmaker size (with manus print width serving as
the size proxy), all of the measures of trackway width
18 J. O. FARLOW ET AL.

Figure 14. Comparison of step lengths between Coffee Hollow A-Male sauropod trackways, and those of sauropods or basal sauro-
podomorphs from other localities. HMNS and HMTHC in the figure key refer to separate measurements made by the Houston
Museum of Natural Science and by the Heritage Museum of the Texas Hill Country, respectively, but the names labelling the track-
ways in the key are HMTHC usage; HMTHC trackway A is the HMNS Right trackway, HMTHC trackway B is the HMNS Middle track-
way, and HMTHC trackway C is the HMNS Left trackway. Other trackway data are from Supplementary Online Material Table 1; all
data are either trackway means or single measurements (when only a single measurement could be made on a trackway). The
Glen Rose Fm trackway label in the figure key is for trackways from that unit consisting of prints of both manus and pes. (A)
Oblique pace length of the manus portion of the trackway vs. manus print width (here and elsewhere, a proxy for trackmaker
size). (B) Stride length of the manus portion of the trackway vs. manus print width. Note that trackway B (Middle trackway) shows
a fairly long pace for its size, but not so much its stride.
 ICHNOS 19

Figure 15. Measures of manus trackway width vs. manus width. Labelling conventions as in Figure 14. All data are trackway
means or single measurements. (A) Inner trackway width. (B, C). Outer trackway width. (B) All trackways. (C) Standard plus manus-
dominant trackways, differentiated on the basis of trackway gauge (narrow-gauge vs. medium- to wide-gauge). (D) Width of angu-
lation pattern. In some comparisons trackway A presents as relatively broad. Trackway B is unusually broad in all relevant
comparisons.

(inner, outer, and WAP) are unsurprisingly correlated
with each other, and more interestingly, with pace
length (Table 3), but either less strongly, or not at all,
with stride length. Because pace as measured here is
oblique to the trackmaker’s direction of travel, this
step length parameter is partly related to the trackway
width to an extent that the stride is not.
Comparisons with other sauropod trackways
(supplementary table 1)
Unsurprisingly, the length of the step (measured as
either the oblique pace or the stride) increases with
increasing sauropod size (Figure 14). Most manus-
only trackways (including Bird’s Mayan Ranch track-
way) and manus-dominant trackways do not differ
from those of standard trackways (those with manus
and pes prints, other than those explicitly identified as
manus-dominant trackways) in relative stride or
oblique pace length. Coffee Hollow trackways A and
C fall within the scatter of standard sauropod track-
way points for both oblique pace and stride, and
trackway B is on the edge of scatter for stride.
Trackway B is different, however, in showing an
unusually long oblique pace for the size of its
manus prints.
All measures of trackway width also unsurprisingly
increase with trackmaker size (Figure 15), although
the relationship between inner trackway width and
manus width shows a great deal of scatter. It is tempt-
ing to attribute this scatter to the existence of wide-,
medium-, and narrow-gauge trackways (Farlow 1992;
Lockley, Farlow, and Meyer 1994; Romano, Whyte,
and Jackson 2007; Marty 2008; Marty et al. 2010).
However, for standard plus manus-dominant track-
ways, analyses of covariance of trackway width against
trackway gauge (with manus print width as the cova-
riate) show only one statistically significant difference
between narrow-gauge trackways on the one hand,
and medium- to wide-gauge trackways, on the other:
that involving outer trackway width (F ¼ 8.059,
p ¼ 0.005, and N ¼ 70 narrow-gauge trackways and
80 medium-wide gauge trackways), and even that
comparison shows very little difference, and considerable
overlap, between the two groups (Figure 15C). As with
step lengths, most manus-only and manus-dominant
20 J. O. FARLOW ET AL.

Figure 16. Pace angulation and manus print rotation as a function of trackmaker size in sauropod and basal sauropodomorph
trackways. (A, B) Pace angulation vs. manus print width. (A) All trackways. (B) Standard plus manus-dominant trackways, differenti-
ated on the basis of trackway gauge. (C, D). Manus rotation vs. manus width. (C) All trackways. (D) Standard trackways. Some
manus-dominant trackways plot within the scatter for standard trackways in the pace angulation vs. manus width comparison, but
two manus-dominant trackways show large pace angulations for the size of their prints. (Data about manus rotation were unavail-
able for trackways identified as manus-dominant.) Manus-only trackways (including the Coffee Hollow trackways) plot within the
scatter for standard trackways in both sets of comparisons; the Coffee Hollow trackways do show pace angulations on the low
side, though.

trackways, including the Mayan Ranch trackway and
Coffee Hollow A-Male trackway C, plot among points
for standard sauropod trackways. Trackway A looks
relatively a bit broader than usual in some compari-
sons, and Trackway B is especially wide in all compari-
sons—not surprising, given that trackway width is
closely related to oblique pace length in sauropodo-
morph trackways (Table 3). However, trackway B is
not unique in its breadth. A standard trackway from
the Late Cretaceous of the southern Pyrenees, attrib-
uted to a titanosaur (Vila et al. 2013: Figure 3(B)), is
also quite broad (Figure 15(A)).
In standard plus manus-dominant sauropodomorph
trackways, the pace angulation shows a weak negative
correlation with manus width (r ¼ 0.160, p ¼ 0.022,
and N ¼ 206; Figure 16(A)), with no difference
between narrow-gauge and medium- to wide-gauge
trackways (Figure 16(B)). Manus-only and manus-
dominant trackways, including those from Coffee
Hollow, plot among the points for standard trackways,
albeit among trackways with pace angulations on the
low side, indicating a less linear, more zig-zagged, track-
way pattern. The Coffee Hollow and Mayan Ranch
trackways have lower pace angulations than typical
sauropod trackways from the Glen Rose Formation.
Manus rotation (positive values ¼ outward rotation,
negative values inward rotation) also weakly decreases
with increasing manus width in standard plus manus-
dominant trackways (r ¼ 0.198, p ¼ 0.036, and
N ¼ 113; Figure 16(C)), again with no difference
between narrow- and medium- to wide-gauge trackways
(Figure 16(D)). Lallensack et al. (2018) also found track-
ways of small sauropods to have surprisingly large
manus rotations (supination angles in their usage). Once
again there is no difference between manus-only and
standard trackways in this comparison. Manus rotation
of trackway B regrettably could not be measured due to
the poor preservation of the individual prints.
The pace angulation, relative trackway width
(expressed as the ratio of trackway width/manus print
 ICHNOS 21

Figure 17. Relative depth of impression of manus and pes prints in sauropod trackways from the Lower Cretaceous of the U.S.
Gulf Coastal plain. (A–C). Glen Rose Formation, Texas. In these trackways pes prints are impressed as deeply as, or more deeply
than, manus prints. (A) Digital model of a negative copy (cast) of a right manus-pes set from the Paluxy River, Dinosaur Valley
State Park, Somervell County (Farlow et al. 2015). Cool colours indicate greater depth. (B, C) Trackways in situ; metre stick provides
the scale. Colour and contrast have been manipulated to enhance distinctness of prints. (B) Slightly oblique view of a right
manus-pes set of a shallowly registered trackway, South San Gabriel River, Williamson County. The dinosaur was moving away
from the viewer, with the metre stick positioned along the trackway midline; pes print length 74 cm. (C) Oblique view of a right
manus-pes set of a very large sauropod, Blanco River, Blanco County. The manus print is 65-70 wide, and the pes print 110 cm
long. (D) LIDAR image of a portion of the Briar Site (De Queen Formation, CertainTeed Gypsum mine, Nashville, Arkansas). Ellipses
surround two sauropod manus-pes sets; width of the manus in the upper set about 53 cm, and width of the manus in the lower
set about 40 cm. The manus prints are slightly more deeply impressed than their associated pes prints (B. F. Platt, personal com-
munication). Note the presence of several manus prints without associated pes prints. Other footprints are attributed to large
theropods. Image courtesy of Platt and the Centre for Advanced Spatial Technologies, University of Arkansas.

width), and the stride/oblique pace ratio are inter-
related (Figure 19). For standard plus manus-domin-
ant sauropod trackways, there is a strong positive
correlation between pace angulation and the stride/
oblique pace ratio (Kendall’s tau-b ¼ 0.784,
Spearman’s rho ¼ 0.912, p for both statistics < 0.001,
and N ¼ 201; Figure 19(B)); both parameters are an
expression of how linear as opposed to zig-zagged a
trackway pattern is. Manus-only and manus-dominant
trackways plot among points for standard trackways,
with Coffee Hollow trackway A (Right trackway) fall-
ing among points at the low end of both variables
(Figure 19(B)); all three Coffee Hollow trackways (and
less convincingly the Mayan Ranch trackway) plot
well below points for typical Glen Rose Formation
sauropod trackways. For standard plus manus-domin-
ant trackways, relative trackway width is at least
weakly negatively correlated with the stride/oblique pace
22 J. O. FARLOW ET AL.

Table 2. Correlations between Heritage Museum of the Texas Hill Country (HMTHC) and Houston Museum of Natural Science
(HMNS) Coffee Hollow A-Male manus print and trackway measurements, and different ways of measuring the same parameter
with HMTHC or HMNS measurements. Data are pooled across trackways; each data point is an individual measurement for a par-
ticular trackway. Significant correlations in bold.
Parameters Pearson’s r p N
Manus width (HMTHC direct) Manus width (HMNS direct) 0.688 <0.001 35
Manus pace (HMTHC print front edge direct) Manus pace (HMNS print front edge diagram) 0.840 <0.001 29
Manus pace (HMTHC print front edge direct) Manus pace (HMTHC print centre map) 0.957 <0.001 47
Manus stride (HMTHC print front edge direct) Manus stride (HMNS print front edge diagram) 0.736 <0.001 23
Pace Angulation (HMTHC print front edge direct) Pace angulation (HMNS print front edge diagram) 0.783 <0.001 22
Trackway inner width (HMTHC direct) Trackway inner width (HMNS diagram) 0.912 <0.001 26
Trackway inner width (HMTHC direct) Trackway inner width (HMNS direct) 0.889 <0.001 27
Trackway inner width (HMNS direct) Trackway inner width (HMNS diagram) 0.978 <0.001 22
Trackway outer width (HMTHC direct) Trackway outer width (HMNS diagram) 0.559 0.003 26
Print rotation (HMNS direct) Print rotation (HMTHC map) 0.225 0.532 10

ratio (inner trackway width/manus width: Kendall’s tau-
b ¼ 0.396, Spearman’s rho ¼ 0.557, and N ¼ 117;
outer trackway width/manus width: Kendall’s tau-b ¼
0.336, Spearman’s rho ¼ 0.481 and N ¼ 153; width
of angulation pattern/manus width: Kendall’s tau-b ¼
0.381, Spearman’s rho ¼ 0.522 and N ¼ 81; p for all
statistics < 0.001). Unsurprisingly, the more linear a
trackway is, the relatively narrower it is. Manus-only
and manus-dominant trackways again plot among
standard trackways, but Coffee Hollow trackway B once
again shows a particularly broad trackway (but not
uniquely so) for its (rather low) values of the stride/pace
ratio Figure 19(D)).
Most standard sauropodomorph trackways show
oblique paces that step off from the left manus to be
about as long as oblique paces that begin with the
right manus (Figure 20). There is a slight suggestion,
albeit one that is not quite statistically significant, that
smaller sauropodomorphs stepped off with longer
paces beginning with the left forefoot, while bigger
sauropods had longer paces beginning with the right
forefoot (standard plus manus-dominant trackways:
Kendall’s tau-b ¼ 0.123, p ¼ 0.073; Spearman’s rho
¼ 0.187, p ¼ 0.063; N ¼ 99 trackways). Most manus-
dominant and manus-only trackways plot among
standard trackways in this relationship. However,
some manus-only sauropod trackways, including
Coffee Hollow trackway A and some trackways from
North Africa (Ishigaki and Matsumoto 2009) show
marked left vs. right stepping off asymmetry.
Forelimb length vs. Manus size in sauropods
There is a great deal of scatter in the limited data
presently available for the relationship between manus
width and forelimb length in sauropods (Figure 21).
Furthermore, both the width of the intact manus, and
the length of the forearm, would have been somewhat
larger in the living animals, due to the presence of
cartilage and other soft tissues, than in skeletal
mounts (cf. Bonnan et al. 2010; Holliday et al. 2010;
Tsai et al. 2018). In any case, the data suggest that the
Coffee Hollow sauropod trackmakers could have had
forelimbs as much as 4 or 5 m in length.
Discussion
Explanations for manus-only sauropod trackways fall
into two categories: preservational or formational arte-
facts in prints created during normal walking (differ-
ential pressure exerted by the manus and pes on the
substrate) or unusual behaviour (‘swimming’). In the
spirit of multiple working hypotheses, we now
consider the pros and cons of each of these interpreta-
tions as applied to the Coffee Hollow A-Male manus-
only sauropod trackways.
Preservation artifact (and perhaps more than one
kind of trackmaker?)
Because differential autopodial pressure plausibly
accounts for most sauropod manus-only trackways
(cf. Lockley and Rice 1990; Vila and Galobart, 2005;
Falkingham et al., 2011a,b; cf. Falkingham, Margetts,
and Manning 2010), this hypothesis must be consid-
ered a potential explanation for the manus-only
nature of the Coffee Hollow trackways. Although
preservation as undertracks probably does not apply
to the Coffee Hollow prints, their very shallowness
elicits suspicion that the absence of pes prints in
these trackways might be due to a formational/pres-
ervational artefact, with relatively greater pressure
exerted on the substrate by the forefeet than by the
hindfeet. As already noted, in presently known
sauropod trackways from the Glen Rose Formation
that consist of both manus and pes prints, whether
faintly/shallowly registered or deeply impressed, the
manus prints are not more deeply impressed than
pes prints (Figure 17). The Briar Site in Arkansas, on
the other hand, preserves associated sauropod
 ICHNOS 23

Figure 18. Comparisons between HMTHC and HMNS measurements. Data cases are individual measurements for the three track-
ways; the same legend applies to all panels. (A) Manus width. (B) Oblique pace. (C) Pace angulation. (D) Trackway inner width. (E)
Trackway outer width. (F) Manus print rotation.

manus-pes sets similar to those from Texas, in a few
of which forefoot prints are deeper than the hindfoot
prints; isolated manus tracks occur nearby.
Consequently unusual depth distributions of manus
and pes prints in sauropod trackways from the
Cretaceous of the U.S. Gulf Coast region could well
be responsible for the unusual occurrence of manus-
only trackways in the Glen Rose Formation. If the
Arkansas trackmakers were all the same kind of
sauropod, this would indicate that under different
circumstances a given sauropod was capable of mak-
ing trackways with very different manus vs. pes
depth distributions.
Another possibility is that the Coffee Hollow and
Mayan Ranch trackmakers were made by a different
kind of sauropod, with a different location of the ani-
mal’s centre of mass, and different pressure loading
on the forefeet and hindfeet, than the makers of most
24 J. O. FARLOW ET AL.

Figure 19. Interrelationships among trackway parameters. Panels A and C plot individual measurements for each of the three
Coffee Hollow trackways, using HMTHC data; panels B and D plot trackway means or single measurements for the Coffee Hollow
(both HMNS and HMTHC data) and other sauropod and basal sauropodomorph trackways. (A, B). The pace angulation and the
stride/oblique pace ratio are positively correlated. (A). Trackway A (Right trackway) shows lower values of both the pace angulation
and the stride/pace ratio than the other two Coffee Hollow trackways. (B). All three Coffee Hollow trackways plot at the low end
of the pace angulation: stride/oblique pace ratio relationship, but among points for standard sauropod trackways. Other manus-
only, and also manus-dominant, trackways similarly plot among the trend defined for standard trackways. (C, D). Trackway inner
width/manus width ratio vs. stride/oblique pace ratio. (C) Each of the three Coffee Hollow trackways plots in a different region of
morphospace, with trackway B showing a particularly broad trackway relative to the stride/pace ratio. (D) Across sauropod track-
ways more generally, relative trackway inner width decreases with increasing values of the stride/oblique pace ratio. Manus-only
and manus-dominant trackways mostly plot among standard trackways. Coffee Hollow trackway B (Middle trackway) and a
‘standard’ trackway of a small sauropod from the Early Cretaceous of China (Chabu 8 D Trackway S4; Lockley et al. 2018) have
unusually broad trackways in this comparison.

other Glen Rose Formation sauropod trackways. At
present, three sauropod genera are known from the
Trinity Group of Texas (D’Emic 2013; cf. Gallup
1989; Tidwell, Carpenter, and Brooks 1999; Wedel,
Cifelli, and Sanders 2000a, b; Rose 2007; D’Emic and
Foreman 2012; Britt et al. 2017): the somphospondyli-
ans Sauroposeidon and Astrophocaudia, and the bra-
chiosaurid Cedarosaurus. Of these three taxa, the pes
is known only for Cedarosaurus, and the manus of all
three is unknown. Farlow, Pittman, and Hawthorne
(1989) assigned the name Brontopodus birdi to sauro-
pod tracks from the Glen Rose Formation of Texas,
with a trackway from R.T. Bird’s Paluxy River track
quarry serving as the type specimen. Although the
morphology of the pes of Cedarosaurus is consistent
with that of hindfoot prints of Brontopodus as dis-
played at the Paluxy River sites (Farlow, Pittman, and
Hawthorne 1989), the same could well have been true
of the pedes of one or both of the other two Trinity
sauropods. On the basis of present knowledge, it is
impossible to say which, if any, of the known Trinity
sauropods was the Brontopodus-maker, and whether
Brontopodus was made by only one sauropod taxon.
Although morphological details of the manus prints
of the three Coffee Hollow A-Male sauropod track-
makers are not particularly clear, their shape seems
consistent with Brontopodus. The manus prints of
Coffee Hollow sauropods A and C are comparable in
size and preservation to the manus prints of the
Blanco River sauropods (Figure 17C), and occur at
the same stratigraphic level, suggesting that they were
produced by the same kind of trackmaker. Pes prints
of the Blanco River trackways, however, are perhaps
different enough in appearance from their Paluxy
 ICHNOS 25

Table 3. Partial correlations between trackway parameters after controlling for overall trackmaker size (in terms of manus width).
Control Variable Trackway Parameters Correlation p Degrees of Freedom
Manus width Stride (HMTHC direct) Trackway inner width 0.172 0.332 32
(HMTHC direct) (HMTHC direct)
Trackway outer width 0.142 0.424 32
(HMTHC direct)
Width of angulation 0.165 0.308 38
pattern (HMTHC)
Trackway inner width Trackway outer width 0.878 <0.001 36
(HMTHC direct) (HMTHC direct)
Width of angulation 0.905 <0.001 30
pattern (HMTHC)
Pace ending in print 0.784 <0.001 33
(HMTHC direct)
Pace beginning with print 0.660 <0.001 35
(HMTHC direct)
Trackway outer Width Width of angulation 0.934 <0.001 30
(HMTHC direct) pattern (HMTHC)
Pace ending in print 0.655 <0.001 33
(HMTHC direct)
Pace beginning with print 0.580 <0.001 35
(HMTHC direct)
Width of angulation Pace ending in print 0.758 <0.001 38
pattern (HMTHC direct)
(HMTHC direct) Pace beginning with print 0.834 <0.001 38
(HMTHC direct)
Manus print width Stride (HMNS diagram) Trackway inner width 0.433 0.021 26
(HMNS Direct) (HMNS diagram)
Trackway inner width 0.427 0.068 17
(HMNS direct)
Trackway outer width 0.432 0.022 26
(HMNS diagram)
Pace ending in print 0.570 0.002 26
(HMNS diagram)
Pace beginning with print 0.630 <0.001 26
(HMNS diagram)
Trackway outer width Pace ending in print 0.746 <0.001 26
(HMNS diagram) (HMNS diagram)
Pace beginning with print 0.753 <0.001 26
(HMNS diagram)
Note: Data are pooled across trackways; each data point is an individual measurement for a particular trackway. Significant partial correlations in bold.

River counterparts that the big south Texas prints
might eventually be assigned to a different ichnospe-
cies than B. birdi (contra Farlow, Pittman, and
Hawthorne 1989). Perhaps one of the Texas sauropods
did have a pedal structure different than that of
Cedarosaurus.
The classic Paluxy River Brontopodus are unques-
tionably true tracks. If the Coffee Hollow and Blanco
River prints also are true tracks, or close to being true
tracks, their makers were likely bigger animals than
most of their north Texas counterparts, which might
further support the conclusion that the south Texas
dinosaurs were a different kind of sauropod.
If the Coffee Hollow trackmakers were different
forms than the Paluxy River trackmakers, the depth
distribution of manus and pes prints seen in the
Paluxy River tracks (Figure 17A) might not also
have characterized the Coffee Hollow sauropods. If
the Coffee Hollow trackways had been made by the
brachiosaurid Cedarosaurus, it would be entirely
expected that the forelimbs exerted more pressure
on the ground than the hindlimbs, potentially
resulting in manus-only trackways (Henderson
2006; Falkingham et al., 2011a). However, keep in
mind that the pes of Cedarosaurus is a good match
for the morphology of the Paluxy River pes prints,
which are consistently as deep as or deeper than
manus prints. The same is true for sauropod track-
ways at other sites in the Glen Rose Formation
(Figure 17(B, C)), including the big Blanco
River prints.
The differential pressure hypothesis for the creation
of Glen Rose Formation manus-only sauropod track-
ways might receive considerable support from discov-
ery of more complete fossils of the three sauropod
taxa known from the skeletal fauna. If it were estab-
lished that one of the sauropod taxa had the morph-
ology and relative sizes of the manus and pes, and a
position of the animal’s centre of mass, consistent
with the production of pes prints as deep or deeper
26 J. O. FARLOW ET AL.

Figure 20. Asymmetry of manus oblique pace lengths in sauropod and basal sauropodomorph trackways. (A) Ratio of the mean
value of paces beginning with the left foot, to the mean value of paces beginning with the right foot. Different symbols are used
for the Coffee Hollow trackways to denote values obtained using different reference points for measuring pace lengths. Most
manus-only or manus-dominant trackways plot among data cases for standard trackways, but Coffee Hollow trackway A and
manus-only trackways from North Africa (Ishigaki and Matsumoto 2009b) show more extreme values. (B) Frequency distribution of
the pace asymmetry ratio in different manus trackway categories. Most trackways in all categories show fairly similar values of
paces beginning with the left, as opposed to the right, manus, but some manus-only trackways show more extreme values.
 ICHNOS 27

Figure 21. Forelimb length vs. manus width in mounted sauropod skeletons. Some of these points should be viewed as very
rough approximations due to extensive reconstruction of missing elements in the skeleton, particularly around the autopodia. Key
to specimens: Amargasaurus (Museo Argentino de Cinencias Naturales ‘Bernardino Rivadavia’, Buenos Aires MACN N-15);
Apatosaurus (Carnegie Museum of Natural History CM 3018); Dicraeosaurus (Museum fu€r Naturkunde Berlin MfN MB.R composite
mount, with majority of material from individual m, and remaining material from individual o [Heinrich 1999]); Diplodocus (Denver
Museum of Nature and Science DMNS 1494); Dreadnoughtus (Museo Padre Molino, Rıo Gallegos, Argentina MPM-PV 1156; 3 D files
from Lacovara et al. 2014—this point is particularly problematic, because the manus is a total reconstruction); Giraffatitan
(Museum fu€r Naturkunde Berlin MfN MB.R mount, primarily based on MfN MB.R.2181); Jobaria (Australian Museum, Sydney MNN
TIG mounted cast of several overlapping individuals, mainly MNN TIG 3); Mamenchisaurus (Institute for Vertebrate Palaeontology
and Palaeoanthropology IVPP multiple specimens, primarily IVPP V456-458); Patagosaurus (cast based on Museo Argentino de
Ciencias Naturales ‘Bernardino Rivadavia’, Buenos Aires PVL 4170); Rapetosaurus (Field Museum of Natural History FMNH PR 2209);
Sauroposeidon (Paluxysaurus) (Fort Worth Museum of Science and History FWMSH 93B-10).

than manus prints (as at the Paluxy River tracksites),
while another sauropod from the Glen Rose
Formation had the autopodial and body proportions
consistent with the production of manus-dominant
trackways, this would tip the scales towards interpret-
ing Glen Rose Formation manus-only sauropod track-
ways as being formational or preservational artefacts.
Even more compelling support for this hypothesis
would be discovery of manus-dominant sauropod
trackways somewhere in the Glen Rose Formation,
with manus prints more deeply impressed than pes
prints, as at the Briar Site in Arkansas. This would
indicate either that a single species of sauropod was
capable of altering its walking to create standard and
manus-dominant trackways as it adjusted to different
substrate conditions, or that there was more than one
kind of footprint-making sauropod in Texas during
the Early Cretaceous, with differences in the location
of the centre of mass and/or the relative sizes of the
front and rear autopodia, between/among the different
trackmakers (Henderson 2006; Falkingham
et al., 2011a).
As previously noted, Coffee Hollow sauropod
trackways A and B are unusually broad for the size of
their makers (Figure 15), but not uniquely so. Vila
et al. (2013: Figure 3(B)) illustrated a standard sauro-
pod trackway from the southern Pyrenees that is simi-
larly broad across both the manus and pes portions of
the trackways. If such a trackway were to be found in
the Glen Rose Formation, and especially if its manus
prints were deeper than its pes prints, this would pro-
vide the strongest possible support for interpreting the
Glen Rose manus-only sauropod trackways as artefacts
of differential autopodial pressure on the substrate.
28 J. O. FARLOW ET AL.

Figure 22. A whimsical exploration of the punting hypothesis for manus-only sauropod trackways (cartoon by R.T. Bakker). The
sauropods are interpreted as brachiosaurids, with longer forelimbs than hindlimbs, unlike R.T. Bird’s interpretation (Figure 1(D)).
The difference in size between the two dinosaurs in the cartoon is probably greater than that between the two large and the
smaller sauropod at Coffee Hollow (Table 1); it is unlikely that a punting sauropod could have been completely submerged, as the
smaller dinosaur in the cartoon is depicted.

Equally or more interestingly, given the differences in
overall size, pace length and trackway width (relative to
manus width), and relative depth distribution of manus
and pes prints between most Glen Rose Formation
sauropod trackways (Figure 17) and this hypothetical
trackway, it would strongly suggest the representation
of more than one kind of track-making sauropod in
the ichnofauna, consistent with the skeletal fauna.
Unusual behavior
Until such a hypothetical trackway is found in the
Glen Rose Formation, though, an alternative hypothesis
for the Coffee Hollow manus-only trackways must be
considered. For at least trackways A (Right) and espe-
cially B (Middle), the trackway pattern differs enough
from what is routinely seen in Glen Rose Formation
sauropod trackways to suggest that these animals might
have been doing something out of the ordinary.
Sauropod B took unusually long oblique paces (Figure
14) and made an unusually broad trackway (Figure 15)
for its size. Sauropod A similarly had a rather broad
trackway (Figure 15), and also showed very marked
asymmetry between steps that began with the left as
opposed to the right forefoot (Figure 20(A)).
Note that if the three Coffee Hollow trackways, con-
trary to our interpretation, were in fact undertracks,
the likely increase in footprint dimensions of under-
tracks relative to true tracks might exacerbate the her-
meneutic problem of the unusual breadth of trackways
A and B, because their makers could have been smaller
animals than their footprint dimensions would other-
wise indicate. Furthermore, any increase in apparent
manus print size due to preservation as undertracks
would also be expected to cause the trackway to appear
narrower than in a corresponding ‘true’ trackway.
Although hypothesized unusual behaviour would
not necessarily involve ‘swimming’, it is worth consid-
ering the possibility that R.T. Bird might have been
correct in thinking that (at least some?) Glen Rose
Formation manus-only sauropod trackways were
made by dinosaurs that were wading in water deep
enough for their makers to punt, pulling themselves
along by their forefeet, while their hind legs floated
above the bottom. Some analyses of the location of
the centre of buoyancy relative to the centre of mass
in certain sauropods (forms related to camarasaurs
and brachiosaurs: Wilson and Fisher 2003; Henderson
2004) suggest that this would cause the hindquarters
of a wading dinosaur to lift off the bottom while the
forefeet were still pushing against the substrate. Such
a punting sauropod might then be able to glide with
its body supported by the water, thereby taking longer
oblique paces and creating a wider trackway than
would be possible for the same animal’s unsupported
body. If much of the dinosaur’s weight was buoyed up
by the water, this might also result in shallowly
impressed manus prints.
On the other hand, any buoyancy/punting hypoth-
esis for the Coffee Hollow manus-only trackways

would require the same centre of mass interpretation
that would result in manus-dominated trackways that
include impressions of both forefeet and hindfeet.
One would expect that a sauropod ‘back-heavy’
enough to be making deeper pes than manus prints
during routine walking would likely punt with its
hindfeet rather than its forefeet. In contrast, it would
be a ‘front-heavy’ walker of the kind likely to produce
manus-dominant trackways that would be the
expected candidate to be a forelimb-punter. On the
face of it, this makes it hard to imagine that the sau-
ropods responsible for the Paluxy River footprints
would have been manus-only punters.
Ironically, then, it might be that discovery of
manus-dominant sauropod trackways in the Glen
Rose Formation would not only make a differential
pressure explanation for manus-only sauropod track-
ways more plausible, but also be a prerequisite for
supporting the hypothesis that manus-only trackways
could have been made by punting sauropods! The
best evidence for forefoot-supported punting sauro-
pods would therefore be discovery of trackways in
which pes prints were absent, but the manus prints
were fairly deeply impressed, indicating a soft sub-
strate in which pes prints would likely have registered,
had the dinosaurs actually been normally walking
rather than punting. This is not true of the shallowly
impressed Coffee Hollow A-Male trackways.
Other than being manus-only, there is nothing
unusual in the pattern of trackway C (Left trackway),
and the same is true of R.T. Bird’s manus-dominated
Mayan Ranch trackway. However, if trackway C was
made at the same time as trackways A and B, and if
the latter two trackways were in fact made by punters,
it would not be unreasonable to think that the same
was true of trackway C.
Although aquatic animals that engage in underwater
pedestrianism may do so while completely submerged
(cf. Zug 1971; Brand 1979; Martinez 1996; Martinez,
Full, and Koehl 1998; Coughlin and Fish 2009; Grigg
and Kirshner 2015; Farlow, Robinson, Turner, et al.
2018a, b), the likely pneumaticity of their bodies makes
it unlikely that sauropods could have done the same
(Henderson 2004). This in turn sets an upper limit to
how deep the water through which hypothetical punt-
ing sauropods moved could have been: probably no
more than a few to several metres (Figs. 21, 22).
Conclusions
The presently available evidence does not decisively
rule out either the differential pressure or the unusual
ICHNOS 29
behaviour hypothesis for the origin of the Coffee
Hollow A-Male manus-only sauropod trackways. On
balance, however, comparison of these trackways with
sauropod trackways from other ichnofaunas provides
stronger support for the differential pressure hypoth-
esis, with the especially interesting, possible corollary
hypothesis that more than one kind of sauropod is
represented in the Glen Rose Formation ichnofauna.
We have described the kinds of trackway evidence
that would allow an unambiguous verdict in favour of
each of the two hypotheses. Now we can hope for
future discoveries in the field to deliver that evidence.
Acknowledgments
This paper is dedicated to the memory of R.T. Bird and
Wann Langston. We commend Ms. Gail Flach and her
associates for their recognition of the importance of the
site, and for interrupting quarry work while study pro-
ceeded. Numerous volunteers assisted in cleaning and docu-
menting the tracksite. For two days, 21 members of the
Heritage Museum of the Texas Hill Country volunteered to
clean, measure, photograph, trace, and map the tracksite.
We thank Matt Bonnan, Don Henderson, Christian Meyer,
Jens Lallensack, and Brian Platt for useful discussions or
unpublished data, and Brian Platt (again) and the
University of Arkansas Center for Advanced Spatial
Technologies for providing the LIDAR image of a portion
of the Briar Site (CertainTeed Gypsum mine). Hendrik
Klein and John Foster made several helpful comments on
the manuscript. Jim Whitcraft assisted in the production of
illustrations.
Disclosure statement
No potential conflict of interest was reported by
the authors.
Funding
This research was supported by a grant from the National
Geographic Society to Farlow. The Department of Library
Services of the American Museum of Natural History kindly
granted permission to reproduce some of R.T. Bird’s images
of the Mayan Ranch site.
References
Alcala, L., F. Perez-Lorente, L. Luque, A. Cobos, R. Royo-
Torres, and L. Mampel. 2014. “Preservation of Dinosaur
Footprints in Shallow Intertidal Deposits of the Jurassic-
Cretaceous Transition in the Iberian Range (Teruel,
Spain).” Ichnos 21 (1):19–31.
Bakker, R.T. 1971. “Ecology of the Brontosaurs.” Nature
229 (5281):172–4.
30 J. O. FARLOW ET AL.
Bates, K.T., P.D. Mannion, P.L. Falkingham, S.L. Brusatte,
J.R. Hutchinson, A. Otero, W.I. Sellers, C. Sullivan, K.
Stevens, and V. Allen. 2016. “Temporal and Phylogenetic
Evolution of the Sauropod Dinosaur Body Plan.” Royal
Society Open Science 3 (3):150636.
Bird, R.T. 1939. “Thunder in his Footsteps”. Natural History
43:254–61.
Bird, R.T. 1941. “A Dinosaur Walks into the Museum.”
Natural History 47:74–81.
Bird, R.T. 1944. “Did Brontosaurus Ever Walk on Land?.”
Natural History 53:61–7.
Bird, R.T. 1954. “We Captured a ‘live’ Brontosaur.”
National Geographic 105:707–22.
Bird, R.T. 1985. Bones for Barnum Brown: Adventures of a
Dinosaur Hunter. Fort Worth, TX: Texas Christian
University Press, 225p.
Bonnan, M.F., J.L. Sandrik, T. Nishiwaki, D.R. Wilhite,
R.M. Elsey, and C. Vittore. 2010. “Calcified Cartilage
Shape in Archosaur Long Bones Reflects Overlying Joint
Shape in Stress-bearing Elements: Implications for
Nonavian Dinosaur Locomotion.” The Anatomical
Record: Advances in Integrative Anatomy and
Evolutionary Biology 293 (12):2044–55.
Brand, L. 1979. “Field and Laboratory Studies on the
Coconino Sandstone (Permian) Vertebrate Footprints
and Their Paleoecological Implications.” Palaeogeography,
Palaeoclimatology, Palaeoecology 28:25–38.
Britt, B.B., R.D. Scheetz, M.F. Whiting, and D.R. Wilhite.
2017. “Moabosaurus utahensis, N. Gen., N. Sp., a New
Sauropod from the Early Cretaceous (Aptian) of North
America.” Contributions from the Museum of
Paleontology, University of Michigan 32:189–243.
Brusatte, S.L., T.J. Challands, D.A. Ross, and M. Wilkinson.
2016. “Sauropod Dinosaur Trackways in a Middle
Jurassic Lagoon on the Isle of Skye, Scotland.” Scottish
Journal of Geology 52 (1):1–9.
Campos-Soto, S.,. A. Cobos, E. Caus, M.I. Benito, L.
Fernandez-Labrador, P. Suarez-Gonzalez, I.E. Quijada, R.
Mas, R. Royo-Torres, and L. Alcala. 2017. “Jurassic
Coastal Park: A Great Diversity of Palaeoenvironments
for the Dinosaurs of the Villar del Arzobispo Formation
(Teruel, eastern Spain).” Palaeogeography,
Palaeoclimatology, Palaeoecology 485:154–77.
Cariou, E., N. Olivier, B. Pittet, J.-M. Mazin, and P.
Hantzpergue. 2014. “Dinosaur Track Record on a
Shallow Carbonate-dominated Ramp (Loulle Section, Late
Jurassic, French Jura).” Facies 60 (1):229–53.
Carvalho, I.D.S., L. Borghi, and G. Leonardi. 2013.
“Preservation of Dinosaur Tracks by Microbial Mats in
the Sousa Basin (Lower Cretaceous), Brazil.” Cretaceous
Research 44:112–21.
Castanera, D., I. DIaz-MartInez, M. Morena-Azanza, J.I.
Canudo, and J.M. Gasca. 2016a. “An Overview of the
Lower Cretaceous Dinosaur Tracksites from the
Mirambel Formation in the Iberian Range (NE Spain).”
In Cretaceous Period: Biotic Diversity and Biogeography,
edited by Khosla, A., and S.G. Lucas, 65–74.
Albuquerque, NM: New Mexico Museum of Natural
History and Science Bulletin 71.
Castanera, D., C. Pascual, J.I. Canudo, N. Hernandez, and
J.L. Barco. 2012. “Ethological Variations in Gauge in
Sauropod Trackways from the Berriasian of Spain.”
Lethaia 45 (4):476–89.
Castanera, D., V.F. Santos, L. Pi~ nuela, C. Pascual, B. Vila,
J.I. Canudo, and J.J. Moratalla. 2016b. “Iberian Sauropod
Tracks Through Time: Variations in Sauropod Manus
and Pes Track Morphologies.” In Dinosaur Tracks: The
Next Steps, edited by Falkingham, P.L., D. Marty, and A.
Richter, 120–37. Bloomington: Indiana University Press.
Colbert, E.H. 1951. The Dinosaur Book: The Ruling Reptiles
and Their Relatives. New York: McGraw-Hill, 156. p.
Console-Gonella, C.,. S. de Valais, R.A. Marquillas, and
M.C. Sanchez. 2017. “The Maastrichtian-Danian Maimara
tracksite (Yacoraite Formation, Salta Group), Quebrada
de Humahuaca, Argentina: Environments and Ichnofacies
Implications.” Palaeogeography, Palaeoclimatology,
Palaeoecology 468:327–50.
Coombs, W.P. Jr. 1975. “Sauropod Habits and Habitats.”
Palaeogeography, Palaeoclimatology, Palaeoecology 17 (1):
1–33.
Coughlin, B., and F.E. Fish. 2009. “Hippopotamus
Underwater Locomotion: Reduced-Gravity Movements
for a Massive Animal.” Journal of Mammalogy 90 (3):
675–9.
Currie, P.J., D. Badamgarav, and E.B. Koppelhus. 2003. The
first Late Cretaceous footprints from the Nemegt locality
in the Gobi and Mongolia. Ichnos 19:1–13.
Dai, H., L. Xing, D. Marty, J. Zhang, W.S. Persons, IV, H.
Hu, and F. Wang. 2015. “Microbially-Induced
Sedimentary Wrinkle Structures and Possible Impact of
Microbial Mats for the Enhanced Preservation of
Dinosaur Tracks from the Lower Cretaceous Jiaguan
Formation Near Qijang (Chongqing, China).” Cretaceous
Research 5:98–109.
Dalla Vecchia, F.M., and A. Tarlao. 2000. “New Dinosaur
Track Sites in the Albian (Early Cretaceous) of the Istrian
peninsula (Croatia) – Part II – Paleontology.” Memorie
de Scienze Geologiche 52 (2):227–92.
D’Emic, M.D. 2013. “Revision of the Sauropod Dinosaurs of
the Lower Cretaceous Trinity Group, southern USA, with
the Description of a New Genus.” Journal of Systematic
Palaeontology 11:707–26.
D’Emic, M.D., and B.Z. Foreman. 2012. “The Beginning of
the Sauropod Dinosaur Hiatus in North America:
Insights from the Lower Cretaceous Cloverly Formation
of Wyoming.” Journal of Vertebrate Paleontology 32:
883–902.
Diedrich, C.G. 2012. “Middle Triassic Chirotherid
Trackways on Earthquake Influenced Intertidal Limulid
Reproduction Flats of the European Germanic Basin
Coasts.” Central European Journal of Geosciences 4:
495–528.
Falkingham, P.L. 2012. “Acquisition of High Resolution
Three-Dimensional Models Using Free, Open-source,
Photogrammetric Software.” Palaeontologia Electronica 15
(1):1T: 15. p.
Falkingham, P.L. 2016. “Applying Objective Methods to
Subjective Track Outlines.” In Dinosaur Tracks: The Next
Steps, edited by Falkingham, P.L., D. Marty, and A.
Richter, 72–80. Bloomington, Indiana: Indiana University
Press.
Falkingham, P.L., K.T. Bates, and J.O. Farlow. 2014.
“Historical Photogrammetry: Bird’s Paluxy River

Dinosaur Chase Sequence Digitally Reconstructed as It
was Prior To Excavation 70 Years Ago.” PLoS One 9 (4):
e93247.
Falkingham, P.L., K.T. Bates, and P.D. Mannion. 2012.
“Temporal and Palaeoenvironmental Distribution of
Manus- and pes-dominated Sauropod Trackways.”
Journal of the Geological Society London 169 (4):365–70.
Falkingham, P.L., L. Margetts, and P.L. Manning. 2010.
“Fossil Vertebrate Tracks As Palaeopenetrometers:
Confounding Effects of Footprint Morphology.” Palaios
25 (6):356–60.
Falkingham, P.L., K.T. Bates, L. Margetts, and P.L.
Manning. 2011a. “Simulating Sauropod Manus-only
Trackway Formation Using Finite-Element Analysis.”
Biology Letters 7 (1):142–5.
Falkingham, P.L., K.T. Bates, L. Margetts, and P.L.
Manning. 2011b. “The ‘Goldilocks’ effect: Preservation
Bias in Vertebrate Track Assemblages.” Journal of the
Royal Society Interface 8 (61):1142–54.
Falkingham, P.L., K.T. Bates, M. Avanzini, M. Bennett, E.M.
Bordy, B.H. Breithaupt, D. Castanera, P. Citton, I. Dıaz-
Martınez, J.O. Farlow, et al. 2018. “A Standard Protocol
for Documenting Modern and Fossil Ichnological Data.”
Palaeontology 61 (4):469–80.
Farlow, J.O. 1992. “Sauropod Tracks and Trackmakers:
Integrating the Ichnological and Skeletal Records.” Zubıa
10:89–138.
Farlow, J.O., J.G. Pittman, and J.M. Hawthorne. 1989.
“Brontopodus birdi, Lower Cretaceous Sauropod
Footprints from the U.S. Gulf Coastal Plain.” In Gillette,
D.D. and Lockley, M.G. (eds.), Dinosaur Tracks and
Traces. Cambridge University Press, Cambridge, 371–94.
Farlow, J.O., N.J. Robinson, M.L. Turner, J. Black, and S.M.
Gatesy. 2018. “Footfall Pattern of a Bottom-walking
Crocodile (Crocodylus acutus).” Palaios 33 (9):406–13.
Farlow, J.O., N.J. Robinson, C.J. Kumagai, F.V. Paladino,
P.L. Falkingham, R.M. Elsey, and A.J. Martin. 2018.
“Trackways of the American Crocodile (Crocodylus acu-
tus) from Northwestern Costa Rica: Implications for
Crocodylian Ichnology.” Ichnos 25 (1):30–5.
Farlow, J.O., W. Langston, Jr., E.E. Deschner, R. Solis, W.
Ward, B.L. Kirkland, S. Hovorka, T.L. Reece, and J.
Whitcraft. 2006. Texas Giants: Dinosaurs of the Heritage
Museum of the Texas Hill Country. Canyon Lake, TX:
Heritage Museum of the Texas Hill Country, 105pp.
Farlow, J.O., K.T. Bates, R.M. Bonem, B.F. Dattilo, P.L.
Falkingham, R. Gildner, J.J. Jacene, G.J. Kuban, A.J.
Martin, M. O’Brien, et al. 2015. “Dinosaur footprints
from the Glen Rose Formation (Paluxy River, Dinosaur
Valley State Park, Somervell County, Texas).” In Early-
and Mid-Cretaceous Archosaur Localities of North-
Central Texas. Field trip guidebook, edited by Noto, C.
75th Annual Meeting, Society of Vertebrate Paleontology,
Dallas, TX, pp. 14–37.
Fondevilla, V., B. Vila, O. Oms, and A.  Galobart. 2017.
“Skin Impressions of the Last European Dinosaurs.”
Geological Magazine 154 (2):393–8.
Gallup, M.R. 1989. “Functional Morphology of the Hindfoot
of the Texas Sauropod Pleurocoelus sp. Indet.” In
Paleobiology of the Dinosaurs, edited by Farlow, J.O.,
71–4. Boulder, CO: Geological Society of America Special
Paper 238.
ICHNOS 31
Grigg, G., and D. Kirshner. 2015. Biology and Evolution of
Crocodylians. Ithaca, NY: Cornell University Press 649p.
Hall, L.E., A.E. Fragomeni, and D.W. Fowler. 2016. “The
Flexion of Sauropod Pedal Unguals and Testing the
Substrate Grip Hypothesis Using the Trackway Fossil
Record.” In Dinosaur Tracks: The Next Steps, edited by
Falkingham, P.L., D. Marty, and A. Richter, 138–50.
Bloomington: Indiana University Press.
Heinrich, W. 1999. “The Taphonomy of Dinosaurs from
the Upper Jurassic of Tendaguru (Tanzania) Based on
Field Sketches of the German Tendaguru Expedition
(1909–1913).” Fossil Record 2 (1):25–61.
Henderson, D.M. 2004. “Tipsy Punters: Sauropod Dinosaur
Pneumaticity, Buoyancy and Aquatic habits. Proceedings
of the Royal Society of London B 271:180–3.
Henderson, D.M. 2006. “Burly Gaits: Centers of Mass,
Stability, and the Trackways of Sauropod Dinosaurs.”
Journal of Vertebrate Paleontology 26 (4):907–21.
Holliday, C.M., R.C. Ridgely, J.C. Sedlmayr, and L.M.
Witmer. 2010. “Cartilaginous Epiphyses in Extant
Archosaurs and Their Implications for Reconstructing
Limb Function in Dinosaurs.” PLoS One 5 (9):e13120.
Huh, M., K.G. Hwang, I.S. Paik, C.H. Chung, and B.S. Kim.
2003. “Dinosaur Tracks from the Cretaceous of South
Korea: Distribution, Occurrences, and Paleobiological
Significance.” The Island Arc 12 (2):132–44.
Hwang, K.-G., M.G. Lockley, M. Huh, and I.S. Paik. 2008.
“A re-interpretation of Dinosaur Footprints with Internal
Ridges from the Upper Cretaceous Uhangri Formation,
Korea.” Palaeogeography, Palaeoclimatology, Palaeoecology
258 (1/2):59–70.
Ishigaki, S. 1988. “Les empreintes de Dinosaures du
Jurasique inferieur du Haut Atlas central marocain.”
Notes du Service Geologique du Maroc 44:79–86.
Ishigaki, S. 1989. “Footprints of Swimming Sauropods from
Morocco.” In Dinosaur Tracks and Traces, edited by
Gillette, D.D., and M.G. Lockley, 83–6. Cambridge:
Cambridge University Press.
Ishigaki, S., and Y. Matsumoto. 2009. “Re-examination of
Manus-only and Manus-dominated Sauropod Trackways
from Morocco.” Geological Quarterly 53:441–8.
Jackson, S.J., M.A. Whyte, and M. Romano. 2009.
“Laboratory-controlled Simulations of Dinosaur
Footprints in sand: A Key to Understanding Vertebrate
Track Formation and Preservation.” Palaios 24 (4):222–8.
Kim, Y.K., K.S. Kim, M.G. Lockley, and S.J. Seo. 2010.
“Dinosaur Skin Impressions from the Cretaceous of
Korea: New Insights into Modes of Preservation.”
Palaeogeography, Palaeoclimatology, Palaeoecology 293 (1-
2):167–74.
Kvale, E.P., G.D. Johnson, D.L. Mickelson, K. Keller, L.C.
Furer, and A.W. Archer. 2001. Middle Jurassic (Bajocian
and Bathonian) dinosaur megatracksites, Bighorn Basin,
Wyoming, U.S.A. Palaios 16 (3):233–54.
Lacovara, K.J., M.C. Lamanna, L.M. Ibiricu, J.C. Poole, E.R.
Schroeter, P.V. Ullmann, K.K. Voegele, Z.M. Boles, A.M.
Carter, E.K. Fowler, et al. 2014. “A Gigantic,
Exceptionally Complete Titanosaurian Sauropod
Dinosaur from Southern Patagonia, Argentina.” Scientific
Reports 4 (1):6196.
Lallensack, J.N., S. Ishigaki, A. Lagnaoui, M. Buchwitz, and
O. Wings. 2018. “Forelimb Orientation and Locomotion
32 J. O. FARLOW ET AL.
of Sauropod Dinosaurs: Insights from the? Middle
Jurassic Tafaytour Tracksites (Argana Basin, Morocco).”
Journal of Vertebrate Paleontology 38 (5):e1512501.
Lee, Y.-N., and M. Huh. 2002. “Manus-only Sauropod
Tracks in the Uhangri Formation (Upper Cretaceous),
Korea and Their Paleobiological Implications.” Journal of
Paleontology 76 (3):558–64.
Lee, Y.-N., and H.-J. Lee. 2006. “A Sauropod Trackway in
Donghae-Myeon, Goseong County, South Korea and Its
Paleobiological Implications of Uhangri manus-only
Sauropod Tracks.” Journal of the Paleontological Society
of Korea 22:1–14.
Lockley, M.G., and A. Rice. 1990. “Did ‘Brontosaurus’ Ever
Swim Out To Sea? Evidence from Brontosaur and Other
Dinosaur Footprints.” Ichnos 1 (2):81–90.
Lockley, M.G., and M.F. dos Santos. 1993. “A Preliminary
Report on Sauropod Trackways from the Avelino Site,
Sesimbra Region, Upper Jurassic Portugal.” Gaia 6:38–42.
Lockley, M.G., J.O. Farlow, and C.A. Meyer. 1994.
“Brontopodus and Parabrontopodus Ichnogen. nov. and
the Significance of Wide- and Narrow-gauge Sauropod
Trackways.” Gaia 10:135–45.
Lockley, M.G., J.C. Garcıa-Ramos, L. Pi~ nuela, and M.
Avanzini. 2008. “A Review of Vertebrate Track
Assemblages from the Late Jurassic of Asturias, Spain
with Comparative Notes on Coeval Ichnofaunas from the
western USA: Implications for Faunal Diversity in
Siliciclastic Facies Assemblages.” Oryctos 8:53–70.
Lockley, M.G., A.P. Hunt, C. Meyer, E.C. Rainforth, and
R.J. Schultz. 1998. “A Survey of Fossil Footprint Sites at
Glen Canyon National Recreation Area (western USA): A
Case Study in Documentation of Trace Fossil Resources
at a National Preserve.” Ichnos 5 (3):177–211.
Lockley, M.G., J. Li, L. Xing, B. Guo, and M. Matsukawa.
2018. “Large Theropod and Small Sauropod Trackmakers
from the Lower Cretaceous Jingchuan Formation, Inner
Mongolia, China.” Cretaceous Research 92:150–67.
Martinez, M.M. 1996. “Issues for Aquatic Pedestrian
Locomotion.” American Zoologist 36 (6):619–27.
Martinez, M.M., R.J. Full, and M.A.R. Koehl. 1998.
“Underwater Punting by an Intertidal Crab: A Novel Gait
Revealed by the Kinematics of Pedestrian Locomotion in
Air versus Water.” Journal of Experimental Biology 201:
2609–23.
Marty, D. 2008. “Sedimentology, Taphonomy, and
Ichnology of Late Jurassic Dinosaur Tracks from the Jura
Carbonate Platform (Chevenez-Ronde tracksite, NW
Switzerland): Insights into the Tidal-Flat
Palaeoenvironment and Dinosaur Diversity, Locomotion,
and Palaeoecology.” Geofocus (University of Fribourg,
Switzerland) 21:113–236.
Marty, D., P.L. Falkingham, and A. Richter. 2016. Dinosaur
track terminology: A glossary of terms. In Falkingham,
P.L., Marty, D., and Richter, A. (eds.), Dinosaur Tracks:
The Next Steps. Indiana University Press, Bloomington,
399–402.
Marty, D., A. Strasser, and C.A. Meyer. 2009. “Formation
and Taphonomy of Human Footprints in Microbial Mats
of Present-day Tidal-flat Environments: Implications for
the Study of Fossil Footprints.” Ichnos 16 (1-2):127–42.
Marty, D., M. Belvedere, C.A. Meyer, P. Mietto, G. Paratte,
C. Lovis, and B. Th€ uring. 2010. “Comparative Analysis of
Late Jurassic Sauropod Trackways from the Jura
Mountains (NW Switzerland) and the Central High Atlas
Mountains (Morocco): Implications for Sauropod
Ichnotaxonomy.” Historical Biology 22 (1-3):109–33.
Mateus, O., and J. Milan. 2010. “A diverse Upper Jurassic
Dinosaur Ichnofauna from Central-west Portugal.”
Lethaia 43 (2):245–57.
Milan, J., and R.G. Bromley. 2006. “True Tracks,
Undertracks and Eroded Tracks, Experimental Work
with Tetrapod Tracks in Laboratory and Field.”
Palaeogeography, Palaeoclimatology, Palaeoecology 231 (3-
4):253–64.
Milan, J., P. Christiansen, and O. Mateus. 2005. “A Three-
Dimensionally Preserved Sauropod Manus Impression
from the Upper Jurassic of Portugal: Implications for
Sauropod Manus Shape and Locomotor Mechanics.”
Kaupia 14:47–52.
Milner, A.R.C., and M.G. Lockley. 2016. “Dinosaur Swim
Track Assemblages: Characteristics, Contexts, and
Ichnofacies Implications.” In Dinosaur Tracks: The Next
Steps, edited by P.L. Falkingham, D. Marty, and A.
Richter, 152–80. Bloomington, IN: Indiana University
Press.
Navarette, R., C.L. Liesa, D. Castanera, A.R. Soria, J.P.
Rodrıguez-Lopez, and J.I. Canudo. 2014. “A Thick
Tethyan Multi-bed Tsunami Deposit Preserving a
Dinosaur Megatracksite Within a Coastal Lagoon
(Barremian, eastern Spain).” Sedimentary Geology 313:
105–27.
Paik, I.S., H.J. Kim, H. Lee, and S. Kim. 2017. “A Large and
Distinct Skin Impression on the Cast of a Sauropod
Footprint from Early Cretaceous Floodplain Deposits,
Korea.” Scientific Reports 7 (1):16339. 7 p.
Pi~
nuela Suarez, L. 2015. Huellas de Dinosaurios y otros
Reptiles del Jurasico Superior de Asturias. Ph. D Thesis,
University of Oviedo, 366pp.
Pittman, J.G. 1989. “Stratigraphy, Lithology, Depositional
Environment, and Track Type of Dinosaur Track-Bearing
Beds of the Gulf Coastal Plain.” In Dinosaur Tracks and
Traces, edited by Gillette, D.D. and M.G. Lockley,
135–153. Cambridge, UK: Cambridge University Press.
Pittman, J.G., and D.D. Gillette. 1989. “The Briar Site: A
new sauropod dinosaur tracksite in Lower Cretaceous
beds of Arkansas, USA.” In Dinosaur Tracks and Traces,
edited by Gillette, D.D. and M.G. Lockley, 311–32.
Cambridge, UK: Cambridge University Press.
Platt, B.F., and S.T. Hasiotis. 2006. “Newly Discovered
Sauropod Dinosaur Tracks with Skin and Foot-pad
Impressions from the Upper Jurassic Morrison
Formation, Bighorn Basin, Wyoming, USA.” Palaios 21
(3):249–61.
Platt, B.F., C.A. Suarez, S.K. Boss, M. Williamson, J.
Cothren, and A.C. Kvamme. 2018. “LIDAR-based
Characterization and Conservation of the First Theropod
Dinosaur Trackways from Arkansas, USA.” PLos One 13
(1):e0190527..
Romano, M., and M.A. Whyte. 2012. “Information on the
foot morphology, pedal skin texture and limb dynamics
of sauropods: Evidence from the ichnological record of
the Middle Jurassic of the Cleveland Basin, Yorkshire,
UK.” Zubıa 30:45–92.

Romano, M., M.A. Whyte, and S.J. Jackson. 2007.
“Trackway Ratio: a New Look at Trackway Gauge in the
Analysis of Quadrupedal Dinosaur Trackways and Its
Implications for Ichnotaxonomy.” Ichnos 14 (3/4):257–70.
Rose, P.J. 2007. “A New Titanosauriform Sauropod
(Dinosauria: Saurischia) from the Early Cretaceous of
Central Texas and Its Phylogenetic Relationships.”
Palaeontologia Electronica 10 (2). http://palaeo-electron-
ica.org/paleo/2007_00063/index.html.
Santos, V.F., M.G. Lockley, C.A. Meyer, J. Carvalho, A.M.
Galopim de Carvalho, and J.J. Moratalla. 1994. A new
sauropod tracksite from the Middle Jurassic of Portugal.
Gaia 10:5–13.
Sanz, E., A. Arcos, C. Pascual, and I. Menendez Pidal. 2016.
“Three-dimensional Elasto-plastic Soil Modelling Analysis
of Sauropod Tracks.” Acta Palaeontological Polonica 61:
387–402.
onberger, J.L., and J.-M. Frahm. 2016. “Structure-from-
Sch€
motion Revisited.” IEEE Conference on Computer Vision
and Pattern Recognition (CVPR): 4104–4113.
Sch€
onberger, J.L., E. Zheng, M. Pollefeys, and J.-L. Frahm.
2016. “Pixelwise View Selection for Unstructured Multi-
view Stereo.” European Conference on Computer Vision
(ECCV), Springer International Publishing, 501–518.
Scott, R.W., A.M. Molineux, H. L€oser, and E.A. Mancini.
2007. “Lower Albian sequence stratigraphy and coral
buildups: Glen Rose Formation, Texas, U.S.A.” In Scott,
R.W. (ed.), Cretaceous Rudists and Carbonate Platforms:
Environmental Feedback. SEPM (Society for Sedimentary
Geology), Special Publication 87, 181–91.
Thulborn, T. 2012. “Impact of Sauropod Dinosaurs on
Lagoonal Substrates in the Broome Sandstone (Lower
Cretaceous), Western Australia.” PLoS One 7 (5):e36208.
Tidwell, V., K. Carpenter, and W. Brooks. 1999. “New
Sauropod from the Lower Cretaceous of Utah, USA.”
Oryctos 2:21–37.
Tsai, H.P., K.M. Middleton, J.R. Hutchinson, and C.M.
Holliday. 2018. “Hip Joint Articular Soft Tissues on Non-
dinosaurian Dinosauromorpha and Early Dinosauria:
Evolutionary and Biomechanical Implications for
Saurischia.” Journal of Vertebrate Paleontology 38 (1):
e1427593.
Vila, B., O. Oms, and A.  Galobart. 2005. “Manus-only
Titanosaurid Trackway from Fumanya (Maastrichtian,
ICHNOS 33
Pyrenees): Further Evidence for an Underprint Origin.”
Lethaia 38 (3):211–8.
Vila, B., O. Oms, A.  Galobart, K.T. Bates, V.M. Egerton,
and P.L. Manning. 2013. “Dynamic Similarity in
Titanosaur Sauropods: Ichnological Evidence from the
Fumanya Dinosaur Tracksite (southern Pyrenees).” PLoS
One 8 (2):e57408.
Ward, W.C., and W.B. Ward. 2007. “Stratigraphy of Middle
Part of Glen Rose Formation (Lower Albian), Canyon
Lake Gorge, central Texas, U.S.A. Lower Albian Sequence
Stratigraphy and Coral Buildups: Glen Rose Formation,
Texas, U.S.A.” In Cretaceous Rudists and Carbonate
Platforms: Environmental Feedback, edited by Scott, R.W..
SEPM (Society for Sedimentary Geology), Special
Publication 87, 193–210.
Wedel, M.J., R.L. Cifelli, and R.K. Sanders. 2000a.
“Osteology, Paleobiology, and Relationships of the
Sauropod Dinosaur Sauroposeidon.” Acta Palaeontologica
Polonica 45:343–88.
Wedel, M.J., R.L. Cifelli, and R.K. Sanders. 2000.
“Sauroposeidon proteles, a New Sauropod from the Early
Cretaceous of Oklahoma.” Journal of Vertebrate
Paleontology 20 (1):109–14.
Whitney, M.I. 1952. “Some Zone Marker Fossils of the Glen
Rose Formation of Central Texas.” Journal of
Paleontology 26:65–73.
Wilson, J., and D. Fisher. 2003. “Are Manus-only Sauropod
Trackways Evidence of Swimming, Sinking, or Wading?”
Journal of Vertebrate Paleontology 23 (3)111A.
Xing, L., G. Peng, M.G. Lockley, Y. Ye, H. Klein, J. Zhang,
and W.S. Persons. IV, 2015a. “Early Cretaceous Sauropod
and Ornithopod Trackways from a Stream Course in
Sichuan Basin, southwest China.” In Fossil Record 4,
Sullivan, R.M., and S.G. Lucas, 319–25. Bulletin 68,
Albuquerque, NM: New Mexico Museum of Natural
History and Science.
Xing, L., Li, D. Falkingham, P.L. Lockley, M.G. Benton, M.J.
Klein, H. Zhang, J. Ran, H. W.S. Persons, IV, and H.
Dai. 2016. “Digit-only Sauropod Pes Trackways from
China—evidence for Swimming or a Preservational
Phenomenon?.” Scientific Reports 6 (1):21138. 11 pp.
Zug, G.R. 1971. “Buoyancy, Locomotion, Morphology of the
Pelvic Girdle and Hindlimb, and Systematics of the
Cryptodiran Turtles.” Miscellaneous Publications of the
Museum of Zoology, University of Michigan, 142, 98p.